There is a setting buried somewhere in almost every PC game released in the last few years called DLSS, or FSR, or XeSS.

Most people leave it on whatever the game sets by default and never think about it again. And that is honestly fine. It works. Your game looks good. Life moves on.

But here is the thing. If you have played any PC game, used a PS5 Pro, or owned a gaming handheld in the last couple of years, your screen has basically been lying to you.

It has almost certainly been showing you something that was never actually created by it. Your graphics card drew some pixels. Software invented the rest.

That is upscaling. And it quietly became one of the most important technologies in gaming.

The reason to talk about it now is because the current AI boom has already made most gaming unaffordable for many.

It’s honestly one of the worst times to be a gamer because RAM prices have shot to the moon, SSD have done the same and modern games become bloated day-by-day.

And upscaling helps in all of this by making your games run better on systems that are aging while being basically free for the most of us.

So it is a good time to start from the beginning and learn what is upscaling, how did we get here, and what is actually happening on your screen right now.

This article is special because it’s part of 3 article series we’ll call ‘Beyond The Frame’.

In this series we’ll focus on gaming technologies that work under-the-hood without us knowing. Technologies like NVIDIA’s DLSS, AMD’s FSR and low latency tech like G-SYNC and V-Sync will be our focus.

The main point is to discuss how they power gaming from consoles to PCs.

The next article will focus on Frame Generation, it’s benefits and limits. We’ll discuss Multi-Frame Gen and try to understand if it’s a good idea to use that or not.

What Even Is Upscaling?

Before any of this makes sense, you need to understand one basic thing about how games work.

When your PC or console renders a video game, it has to basically draw every single frame you see. At 60 frames per second or 60 Hertz, it is drawing 60 complete images every second.

Each one of those images has to have every object in the scene calculated, lit, and placed correctly. That is a lot of work.

The higher the resolution you play at, the more pixels need to be drawn per frame. Running a game at 4K means drawing roughly four times as many pixels as 1080p. For a GPU trying to hit smooth frame rates, that adds up fast.

If you’re not sure what the words “resolution” or “hertz” mean here, you might want to check our Monitor Buying article where we talked about those in more detail, that article is linked right here.

Upscaling is the answer the industry landed on. Instead of drawing the game at full 4K, your GPU draws it at a lower resolution, say 1080p or 1440p.

Upscaling software then takes that lower-resolution image and blows it up to 4K before it appears on your screen.

The result is that your GPU does far less work per frame. Games run faster, or can look much better, without needing a significantly more powerful graphics card.

This is the gist of upscaling in video games. You run your game at a lower resolution than your native resolution, this makes the game run smoother at first. Then AI takes over and upgrades those frames to a higher resolution.

That is basically what we need to know as gamers when it comes to upscaling. Of course there’s a lot more to it but I won’t bore you by converting this article into a research paper :p

How AI Upscaling Works?

You might be wondering what the catch is. How can the graphic card just take a normal resolution image and make it better?

Well, we all know that if you zoom into a digital photo, it gets blurry and pixelated the more you zoom into it. Upscaling in a way is doing that too.

That begs the question: Why does upscaling not give the blur effect?

The old way of upscaling, which games used for years, did look blurry. It was called bilinear or bicubic upscaling, and it basically averaged pixel values to fill in the missing ones.

The result was softer than native resolution and often quite noticeable. Modern upscaling however, is different because it uses AI.

Modern upscaling software is trained on thousands and thousands of high-resolution game images. Over time it learns how edges, textures, hair, grass, and other game elements look at full resolution.

When it sees a lower-resolution image, it does not just average pixels. It looks at the scene and intelligently guesses what the missing detail should look like, based on what it has learned.

Modern upscalers like DLSS and FSR all provide AI-based upscaling and they are genuinely hard to tell apart from native resolution, even to experienced eyes looking closely.

That is not just marketing. Blind tests have repeatedly confirmed it.

If you don’t have two computers running side-by-side with one using upscaling and the other without it, you probably won’t figure out if AI upscaling is being used in a video game or not.

How NVIDIA’s DLSS Became The Gold Standard

DLSS stands for Deep Learning Super Sampling. This is NVIDIA’s upscaling technology that was launched in 2018 alongside their RTX 20-series cards.

When it comes to upscaling technology inside PC gaming, DLSS is the industry standard. But that wasn’t the case at its launch.

The first version of DLSS was not impressive. Most reviewers said native resolution still looked clearly better and the performance gains were not worth the visual trade-off.

And this was the main thing, reviewers were able to tell the difference just by looking at the image.

DLSS 2 got to us in 2020. It was the real moment. NVIDIA rebuilt the underlying AI model completely, and reviewers who had dismissed the first version changed their minds quickly.

At quality settings, DLSS 2 output was genuinely difficult to distinguish from native resolution in most games.

What makes DLSS different from every other upscaler is that it requires dedicated hardware inside NVIDIA’s RTX graphics cards called tensor cores.

These are chips specifically designed to run AI workloads fast. That is why DLSS only works on NVIDIA RTX cards. There is no version for AMD or Intel GPUs.

The trade-off is that NVIDIA can train a much more sophisticated model because it knows exactly what hardware will be running it. The quality shows.

DLSS 3 and 4 built on DLSS 2 supporting an even bigger catalog of games and gradually fixing much of what was wrong with DLSS 2.

At CES 2026, NVIDIA announced DLSS 4.5, which introduced their second-generation AI model. The new model was trained on a significantly larger dataset and uses five times the compute of the previous version.

The result is better stability, cleaner edges, and improved motion clarity across all RTX cards.

DLSS 4.5 on RTX 50-series cards also brought a 6x Multi Frame Generation mode, which I will explain more in the frame generation section.

AMD’s Answer To DLSS

FSR stands for FidelityFX Super Resolution. AMD launched it in 2021 as an answer to NVIDIA’s DLSS.

The big pitch from day one was that FSR works on any GPU. FSR has been open-source which means that no matter if you have an NVIDIA, AMD, Intel card, you can just use FSR as long as the game supports it.

FSR does not require special AI hardware because it is built on algorithms that run on the regular shader cores every GPU has.

That openness was genuinely valuable. Anyone who had a graphics card could turn FSR on and get more performance out of it.

The trade-off was image quality. Early versions of FSR were noticeably softer and shakier than DLSS, especially in motion.

When a game had lots of small moving objects, or fine detail like hair or chain fences, FSR tended to struggle while DLSS held up well.

FSR 4, released in 2025, changed that significantly. AMD finally adopted an AI-powered approach similar to DLSS, and the jump in quality was substantial.

Blind tests confirmed that FSR 4 finally brought AMD’s upscaling into competitive territory with NVIDIA.

The catch is that FSR 4’s best quality requires AMD’s new RDNA 4 cards, like the RX 9070 series.

The older broad compatibility that made FSR famous does not fully apply to FSR 4’s highest-quality mode. Older AMD cards, and NVIDIA and Intel cards, still get the older FSR versions.

In March 2026, AMD also released FSR 4.1, which is based on the same neural network as Sony's updated PSSR. I'll talk about that in a moment.

FSR being competitive is one of the big reasons behind AMD’s comeback. If you want to know how, you can checkout the article linked here where we went deep into AMD’s failures and successes in graphics and computing.

Intel Joining The Upscaling Bandwagon

XeSS stands for Xe Super Sampling. It is Intel’s upscaling technology, released alongside their Arc graphics cards.

It does not get much attention because it is the new kid on the block when it comes to upscaling tech on PC graphics cards.

Like FSR, XeSS works on any GPU. But when running on Intel Arc cards, it uses dedicated AI acceleration hardware on those chips to produce better results than the standard version.

Think of it as a middle ground between DLSS and FSR in terms of how it scales with hardware. It is not totally exclusive like DLSS and not open-source like FSR. You can run it on other GPUs but it works best on Intel GPUs.

Image quality on XeSS sits between older FSR and DLSS in most comparisons. Not quite at DLSS levels, but better than FSR 3.

In 2025, Intel released XeSS 2.1, which quietly expanded frame generation support to NVIDIA and AMD cards as well, not just Intel hardware.

That makes Intel the most broadly compatible option in the room right now for both upscaling and frame generation across different GPU brands.

It does not get the headlines DLSS gets. But if you are on a non-NVIDIA, non-AMD card, or simply want the widest compatibility, XeSS is worth knowing about.

PlayStation Uses Upscaling Too

Consoles used to rely on fairly basic upscaling. The PS4 and base PS5 used AMD’s FSR in many games, which as we covered was not always flattering.

Sony decided to build their own.

PSSR stands for PlayStation Spectral Super Resolution. It launched with the PS5 Pro in late 2024.

Sony built PSSR as part of a joint project with AMD called Project Amethyst. The collaboration let Sony tune an AI upscaling model specifically for the PS5 Pro’s hardware and how it renders games.

The original PSSR was well-received. PS5 Pro games using it looked noticeably sharper and cleaner than the same games on the base PS5 using FSR.

Then in March 2026, an updated version of PSSR launched alongside Resident Evil Requiem. This version, often called PSSR 2.0, is based on the same neural network as AMD’s FSR 4.1.

PS5 architect Mark Cerny confirmed in an interview that the two share the same core technology, just trained differently to suit the different hardware in the console versus a PC GPU.

Reviewers who tested the updated PSSR said it is now very close to DLSS quality. For PS5 Pro owners, that is a meaningful upgrade.

In a way, PSSR and FSR 4.1 are like cousins. They have the same underlying neural network but are trained differently for different hardware.

Frame Generation

Upscaling makes your image look sharper. Frame generation does something different. It makes your frame rate appear higher.

Here is how it works. Your GPU renders one real frame. Frame generation software then invents additional frames between each real one and inserts them into the sequence your monitor displays.

With DLSS 4.5’s 6x Multi Frame Generation on RTX 50-series cards, your GPU renders one frame and the software generates up to five more around it before display.

The frame rate shown on screen can be dramatically higher than what your GPU is actually producing.

This is where the “23 out of 24 frames are AI” number you might have seen comes from. At 6x frame generation, only one in six frames is real.

The obvious question is whether those generated frames look and feel real. For the image quality side, they largely do in slow-paced or cinematic games. The bigger issue is input lag.

Generated frames are invented after the fact. They do not know what you did with your controller or mouse in the moment they represent. Real frames do.

So frame generation adds a small delay between your input and what you see on screen.

For a story-based game or an open-world exploration game, that delay is barely noticeable. For competitive multiplayer, especially at competitive skill levels, it matters.

Most professional or competitive players turn frame generation off.

I will cover frame generation in much more depth in a separate article, because it genuinely deserves its own piece.

For now, the short version is that Frame Generation works but it has real trade-offs, and you should understand it separately from upscaling even when companies sometimes mix them up.

DLSS 5 and the AI slop Controversy

A couple of months ago NVIDIA took the stage at their GTC keynote and announced DLSS 5.

This is not upscaling. This is something different even though they call it DLSS.

DLSS 5 uses generative AI (think of tech behind ChatGPT) to actively alter how games look in real time, not just sharpen a lower-resolution image.

It analyses the scene and applies what NVIDIA described as neural rendering, re-drawing things like skin texture, lighting, and facial detail to look more real.

The first game they showed was Resident Evil Requiem, one of the most visually praised games in recent memory.

The demo showed the main character Grace Ashcroft with smoother skin, removed eye bags, and softened facial features. The internet called it “yassified” immediately.

It looked like an Instagram beauty filter applied to a character that was already beautifully designed. The YouTube reveal was heavily disliked.

Then things got more complicated. It was reported that developers at Capcom and Ubisoft found out about the demo at the same time as the public.

One Ubisoft developer was quoted saying “We found out at the same time as the public.”

Jensen Huang, NVIDIA’s CEO, said in an interview the day after the reveal that critics were “completely wrong” because the technology gives developers full control over the result.

Then, a few days later, on the Lex Fridman podcast, his tone shifted. “I think their perspective makes sense and I can see where they’re coming from,” he said. “I don’t love AI slop myself.”

The important context is that DLSS 5 is not yet available. It launches in fall 2026, only on RTX 50-series cards.

When it comes to DLSS 5, everything shown up till now was a controlled demo. We can only know for sure when it gets released later this year.

How Upscaling Is Helping Gamers Like Me

Upscaling went from a niche setting you turned on if you wanted a few extra frames to something that is now assumed by default.

Chances are that if you have a compatible GPU and a supported game you’re already using upscaling technologies to play your games.

Game developers increasingly design and test their games with upscaling turned on because that’s where consumers are at. It is no longer an optional setting for modern gaming.

New gaming hardware that would come out (like Valve’s Steam Machine) would actively use the upscaling tech from AMD to provide a promised 4k 60Hz experience.

DLSS 4.5 and FSR 4 both produce results that are hard to distinguish from native resolution in most scenarios at least for the average Bashir.

And especially now when AI has jacked prices for most components that make up our gaming consoles and gaming PCs, it’s high time that we get better upscaling technologies.

I use an AMD R9 290, it’s an old graphics card from 2013 and guess what? I can use FSR on it. Isn’t it great that I can still suck some juice out of this old beast even when it has given a decade of service?

A couple of months ago I wanted to play Uncharted 4 on this same graphics card. And it was really struggling to launch the game.

I was hardly getting frames in the 30s I guess. But turning on FSR made the experience so great that I was able to play the game.

I didn’t finish the game but the point is that I had the option to play it on my old hardware just because companies like NVIDIA and AMD invested their time and effort in upscaling technologies.

With PC prices rising globally especially around this AI hype cycle that we’re going through, upscaling technologies are what gamers like me rely on.

And while generally upscaling technologies are great, NVIDIA just showed us in their DLSS 5 demo that things can really get out of hand too.

Whether NVIDIA ships DLSS 5 the way it is this fall is to be seen honestly. The people have shown their frustration, let’s see how NVIDIA takes that.

For now, the simplest version of all of this is that your GPU has likely been lying to you all this while. But the pixels it is inventing are getting very good.

And that’s breathing some more life to our aging gaming hardware.

This article is special because it’s the first in a series of three articles on background graphic technologies in video games.

The next article would go deep into Frame Generation technologies from NVIDIA, AMD and Intel.

Please leave any feedback for this article or the series, in the comments down below:

Further Learning

• What Is AI Upscaling?

• What Is DLSS? Demystifying Nvidia’s Deep Learning Supersampling Technology

• What Is AMD FSR? FidelityFX Super Resolution Explained

• NVIDIA DLSS 4.5 announcement

• Gamers Hate Nvidia’s DLSS 5

• “I Don’t Love AI Slop Myself” – Jensen Huang

• Nvidia CEO Jensen Huang backtracks

• Capcom and Ubisoft devs were out of the loop on Nvidia DLSS 5

• AMD locked FSR 4 behind new GPUs

• DLSS vs FSR vs XeSS Support Across 650+ Games

• Intel Opens Up XeSS 2 to NVIDIA and AMD GPUs

• PlayStation 5’s PSSR 2.0 uses AMD’s FSR 4

• Nvidia says DLSS 4 looks better than native resolution

• AMD FSR 4 vs. Nvidia DLSS 4: Shockingly close

In 2015 AMD’s stock was less than 2 dollars per share.

The company was bleeding money. Their CPUs were getting destroyed by Intel. Their GPUs were losing to Nvidia. There were talks of bankruptcy.

Fast forward to today. AMD’s stock is near 450 dollars. They have arguably the best consumer CPUs.

AMD controls 30 percent of the consumer CPU market. Their server chips are taking share from Intel in data centers too.

Their GPUs still trail Nvidia but the gap is closing. The latest version of AMD’s upscaling technology is pretty good for most of us out here and FSR5 is going to be even more exciting.

People used to say AMD GPUs are not stable enough but today every major gaming console from Playstation, XBOX to handhelds like the Steam Deck, all use AMD GPUs.

And the next generation of consoles are also going to use AMD. Valve has confirmed that for their upcoming Steam Machine.

Meanwhile PlayStation and XBOX have leaks that they’re working with AMD for their next consoles.

This is the story of how AMD refused to die despite being smaller, poorer, and less powerful than both of their competitors.

It is also the story of one of the biggest comebacks in tech history. And why that comeback still might not be enough.

AMD Before Lisa Su

AMD’s story of the last decade revolves around her current CEO who took charge back in 2014.

She has been so instrumental in the company’s success that I think it’s best to see AMD as the one before she took over and after she became CEO.

AMD was founded in 1969 by Jerry Sanders.

Sanders was a former Fairchild Semiconductor executive who got pushed out in a corporate shakeup.

The company started in Sanders’ living room in Santa Clara with eight employees. The plan was simple. Make copies of other companies’ chips under license.

Back in the 1970s big companies did not want to depend on a single chip supplier. So AMD became a second source manufacturer.

They made Intel 8080 processors. They made more unknown chips. They made whatever customers needed them to make.

This worked for a while. AMD went public in 1972. In 7 years, their revenue hit 100 million dollars.

But they were always a follower, never a leader and Sanders knew this.

In a famous quote Sanders said “Real men have fabs” meaning that serious chip companies needed to own their own manufacturing.

So AMD built fabs. They invested in R&D. And in the 1990s they started designing their own x86 processors.

The K5 launched in 1996. It was AMD’s first internally designed CPU. It was not very good. Intel’s Pentium was faster.

The K6 in 1997 was better. AMD partnered with NexGen whose technology formed the basis of the K6. It competed well with Intel’s Pentium II in the budget segment.

But the real breakthrough came in 1999 with the Athlon.

Athlon Joins The Chat

The Athlon was the first x86 processor to hit 1 GHz. AMD beat Intel to that milestone by a few months.

More importantly, the Athlon was genuinely fast. In many benchmarks it matched or beat Intel’s Pentium III. Suddenly AMD was not just the budget option. They were competitive.

AMD’s revenue grew to 2.8 billion dollars in 1999. Then 4.6 billion in 2000. The company was on a roll.

Then came the Athlon 64 in 2003.

The Athlon 64 was AMD’s most important chip ever. It introduced 64-bit computing to the x86 platform.

Intel was pushing Itanium as their 64-bit solution. Itanium was not x86 compatible. It required all new software.

Nobody wanted that. The industry wanted 64-bit extensions to x86 so old software would still work.

AMD delivered exactly that with AMD64. Intel eventually had to license AMD’s 64-bit extensions and rebrand them as Intel 64.

From 2003 to 2006 AMD had the better product. The Athlon 64 used less power than Intel’s Pentium 4. It ran cooler. And in many workloads it was faster.

Intel’s Pentium 4 architecture was deeply flawed. Intel had bet everything on clock speed. The Pentium 4 hit 3.8 GHz but it achieved that speed through a very long pipeline.

The Pentium 4 was fast on paper but slow in practice for many tasks. The Pentium 4 also consumed massive amounts of power.

Intel’s Pentium 4 could hit 115 watts at load. That is more than many GPUs.

AMD’s Athlon 64 consumed around 90 watts and was often faster. Enthusiasts and gamers loved it.

AMD’s market share hit 25.3 percent in Q4 2006 according to Mercury Research. That was AMD’s peak.

Then Intel struck back.

Intel Strikes Back With Core 2 Duo

In July 2006 Intel launched the Core 2 Duo. This was a completely new architecture. Intel had basically admitted the Pentium 4 was a mistake and started over.

The Core 2 Duo was everything the Pentium 4 was not. It was power efficient. It had excellent performance per clock. It scaled well with more cores.

And it destroyed AMD’s Athlon 64 in benchmarks.

By late 2006 Intel had regained the performance crown. AMD’s market share started dropping. Revenue peaked at 5.6 billion in 2006 then declined.

Intel also had way more money to spend. In 2006 Intel’s R&D budget was 5.9 billion dollars. AMD’s entire revenue was 5.6 billion.

AMD could not compete with Intel’s spending power. They needed something to change the game.

The ATI Acquisition

In July 2006 AMD announced they were buying ATI Technologies for 5.4 billion dollars.

ATI was Nvidia’s main competitor in graphics cards. They made the Radeon GPU series. They were roughly equal with Nvidia in market share at the time.

AMD’s vision was to combine CPUs and GPUs on the same chip. They called this an APU or Accelerated Processing Unit.

The idea made sense. Graphics were becoming more important for computers. Games needed GPUs. Video playback benefited from GPU acceleration.

If AMD could integrate a decent GPU with their CPU they could offer better value than Intel whose integrated graphics were terrible.

But the acquisition was a disaster from a financial perspective.

AMD paid 5.4 billion dollars. They paid 2.5 billion in cash and took on 2.5 billion in debt. The remaining 400 million was in stock.

This was AMD’s entire cash reserve plus massive new debt. And the timing was terrible.

2008 Financial Crisis

The 2008 financial crisis hit. PC sales collapsed. AMD’s revenue dropped from 6.0 billion in 2007 to 5.8 billion in 2008 to 5.4 billion in 2009.

AMD was burning cash. Their debt was ballooning. They had to pay interest on billions of dollars while losing money too.

In 2009 AMD lost 1.2 billion dollars. In 2011 they lost 1.8 billion. In 2012 they lost 1.2 billion again.

The company was slowly dying.

To survive AMD had to sell assets. The biggest sale was their chip manufacturing business.

In 2009 AMD spun off their fabs into a separate company called GlobalFoundries. The deal was structured as a joint venture with a company from Abu Dhabi.

AMD got 700 million dollars immediately and another 500 million later. This gave AMD cash to survive. But it also meant AMD no longer controlled their own manufacturing.

AMD went from being an integrated device manufacturer like Intel to being fabless like Nvidia and Qualcomm.

They had to rely on GlobalFoundries to make their chips. From the looks of it, they lost the ability to optimize manufacturing for their designs.

However, this is something that is going to flip the table as you read on.

Meanwhile on the product side things were getting worse. AMD’s CPU designs were falling further behind Intel. And their GPU designs were competitive but not winning against Nvidia.

AMD needed a new CPU architecture. Something to leapfrog Intel. They bet everything on a new CPU architecture from the ground up.

The CPU Architecture That Nearly Killed AMD

Bulldozer was supposed to save AMD. Instead it nearly killed them. The irony lol.

The architecture launched in October 2011 with the FX-8150 processor. AMD marketed it as the world’s first 8-core desktop CPU.

The problem was it was not really 8 cores. Not in the way most people understood cores.

Bulldozer used a module design. Each module had two cores but they shared a single floating point unit and they shared other resources.

AMD counted this as two cores. Technically they were right. But in practice two Bulldozer cores in a module performed more like 1.5 traditional cores.

Intel’s Core i7 2600K from early 2011 had 4 real cores. It destroyed the 8-core FX-8150 in most benchmarks.

In single-threaded tasks Bulldozer was terrible. The FX-8150 was slower than AMD’s own Phenom II chips from 2009.

Gaming performance was bad because games in 2011 mostly used 1-4 threads. Bulldozer’s theoretical 8 cores did not help.

Power consumption was high. The FX-8150 had a 125 watt TDP. Intel’s i7 2600K was 95 watts and faster.

Reviews were brutal too. AnandTech called it “disappointing.” Tom’s Hardware said “we cannot recommend Bulldozer.”

AMD’s stock dropped 30 percent in the months after Bulldozer launched.

But AMD doubled down on the architecture. They released Piledriver in 2012, Steamroller in 2013, and Excavator in 2015.

All of them were based on the same flawed Bulldozer design. Minor improvements each generation but the fundamental problems remained.

For five years from 2011 to 2016 AMD had no competitive CPU in the enthusiast or high-end market.

Their market share collapsed. In Q3 2015 AMD had just 17.7 percent of the x86 CPU market. Intel had about 82.3 percent.

On top of the Bulldozer disaster AMD faced a class action lawsuit.

Customers sued AMD claiming the FX processors were falsely advertised. The lawsuit argued that Bulldozer’s modules were not true cores.

AMD fought the lawsuit for years. In 2019 they settled for 12.1 million dollars without admitting wrongdoing.

But the reputation damage was done. Enthusiasts and gamers avoided AMD CPUs from there on and tech reviewers recommended Intel by default.

That’s partly where the GPU stability and excessive heat reputation comes from, for AMD. Ask an AMD user from that time and they will tell you about it.

AMD’s CPU business was dying and on the GPU side things were not much better.

Did you know? Intel quite recently had its own Bulldozer moment. Checkout this linked article for complete story of how Intel lost its processor crown.

Struggling GPU Business

AMD’s Radeon HD 7000 series in 2012 was competitive with Nvidia’s GTX 600 series.

Performance was similar but Nvidia had advantages beyond raw performance.

Nvidia’s drivers were more stable. Game developers optimized for Nvidia cards because that is what most gamers bought.

Nvidia had PhysX for physics simulation in games. AMD had nothing comparable.

Most importantly, Nvidia had CUDA.

CUDA was Nvidia’s programming platform for general purpose GPU computing. It launched in 2006 and by 2012 it was the standard for scientific computing on GPUs.

Researchers used CUDA for physics simulations, molecular dynamics, climate modeling, and early machine learning.

CUDA has been instrumental to Nvidia’s 5 Trillion dollar market cap. You can check this linked article for our in-depth coverage of Nvidia.

AMD’s equivalent was OpenCL. OpenCL was open source and theoretically better because it worked on any hardware.

But CUDA had a five year head start. The ecosystem was built around CUDA. Universities taught CUDA. Researchers published CUDA code.

AMD could not overcome that momentum and by 2015 AMD was in crisis on both CPUs and GPUs.

Their revenue had dropped to 4.0 billion dollars. They lost 660 million dollars that year. Their stock was at 1.61 dollars.

AMD was basically cooked to perfection :p

The Year AMD Almost Died

In January 2015 AMD’s market cap was just 1.5 billion dollars. Intel’s market cap was 160 billion. Nvidia’s was 11 billion.

AMD was tiny. And they were running out of money. The company had 785 million in cash at the end of 2014. They were burning through it quickly.

AMD sold their stake in a Chinese joint venture for 293 million dollars. They negotiated a new 2.25 billion dollar debt which bought AMD some more time.

The consensus was that AMD would either go bankrupt or get bought by a larger company. Samsung and Qualcomm were mentioned as possible options.

AMD needed new products. But new products take years to develop. And AMD did not have years of cash runway.

The person who had to solve this impossible problem was Lisa Su.

Lisa Su Takes Charge

Lisa Su initially joined AMD in 2012 as senior vice president of global business units. She was promoted to COO in 2014 and became CEO in October 2014.

Su had serious credentials. She earned a PhD in electrical engineering from MIT. She worked at Texas Instruments, IBM, and Freescale Semiconductor before joining AMD.

At IBM she worked on the Cell processor used in the PlayStation 3. She understood chip design at a deep level.

When Su became CEO AMD was losing 180 million dollars per quarter. The product pipeline was a mess. Employee morale was terrible.

Su made hard decisions fast.

First she killed unprofitable product lines. AMD had been trying to make chips for tablets and smartphones. They were failing badly.

Su shut down those programs. She cut 5 percent of AMD’s workforce. She focused everything on three areas: gaming, data centers, and embedded chips.

Second she bet on a completely new CPU architecture.

AMD had a team working on a clean-sheet design codenamed Zen. This was not a Bulldozer derivative. It was a fresh start.

Zen has been in development since 2012 under architect Jim Keller. Keller was a legend in chip design. He had worked on the original Athlon and the Athlon 64.

Keller left AMD in 2015 but the Zen team continued under Mike Clark and Suzanne Plummer.

Su committed AMD’s limited resources to Zen. If it failed AMD was probably done.

But Zen had a secret weapon that could make AMD competitive even without Intel’s manufacturing advantage - Chiplets.

The Ryzen Gamble

Traditional high-performance CPUs are monolithic which means that the entire processor is carved from one piece of silicon.

Want an 8-core CPU? You need a piece of silicon large enough for 8 cores plus cache plus memory controllers plus IO.

The bigger the silicon die the more expensive it is. And bigger dies have a worse problem: yield.

Yield is the percentage of chips that work after manufacturing. Not every chip on a wafer works. Some have defects.

The bigger the die the more likely it contains a defect. This is why Intel’s massive 28-core Xeon chips cost 3000 to 10000 dollars. The dies are enormous and the yields are terrible.

AMD could not afford to make giant monolithic chips because they did not own their own fabs. They had to pay TSMC or GlobalFoundries per wafer.

Low yields would make AMD’s chips too expensive to compete. So AMD’s engineers invented a different approach. Chiplets.

Instead of one big die you make multiple small dies. Each small die has a few CPU cores and cache. Then you connect the small dies together using a high-speed interconnect.

Small dies have better yields. And you can mix and match chiplets to make different products.

Need a 6-core CPU? Use one chiplet. Need a 16-core CPU? Use more chiplets.

This was revolutionary. No major CPU company had done this successfully before.

The challenge was the interconnect. If the connection between chiplets was too slow the CPU cores could not communicate efficiently. Performance would tank.

AMD spent years developing Infinity Fabric. This was their chip-to-chip interconnect technology.

Infinity Fabric had to be fast enough that cores on different chiplets could share cache and memory without huge latency penalties.

It also had to use reasonable amounts of power. The interconnect could not burn watts just moving data between chips.

AMD bet the company that Infinity Fabric would work.

The first Zen chips launched in March 2017 as Ryzen 7.

The flagship Ryzen 7 1800X had 8 cores and 16 threads. It cost 499 dollars.

Intel’s equivalent the Core i7-6900K had 8 cores and cost 1089 dollars.

Ryzen offered 90 percent of the performance for less than half the price.

Reviews were positive. Tom’s Hardware said Ryzen was “competitive with Intel’s best.” PC Gamer called it “the CPU competition we’ve been waiting for.”

But Ryzen was not perfect. Gaming performance was weaker than Intel because of Infinity Fabric latency and lower clock speeds.

Still Ryzen sold well. Enthusiasts who wanted high core counts for content creation and streaming bought Ryzen in huge numbers.

AMD’s CPU market share started climbing. From 17.7 percent in Q3 2015 to 19.6 percent in Q4 2017.

But the real target was not gamers. It was the data center.

Beating Intel At Their Own Game

Intel made most of their profit from Xeon server CPUs.

Data centers bought Xeon chips by the thousands. Prices ranged from 500 dollars to 10000 dollars per chip.

Intel’s data center group generated 19.1 billion dollars in revenue in 2016. That was 33 percent of Intel’s total revenue.

Intel’s Xeon chips are expensive and meant for enterprise but when bought used they can be excellent value for consumers.

Linked here is my article that shows how used Xeon processors can provide you a working computer for close to $250.

AMD’s server CPU business was basically dead. Their Opteron chips based on Bulldozer were terrible. AMD had less than 1 percent server market share.

But with Zen AMD had a chance.

In June 2017 AMD launched EPYC for servers. These chips used up to four 8-core chiplets for a total of 32 cores.

Intel’s Xeon processors topped out at 28 cores and cost more than equivalent EPYC chips.

At first data centers were skeptical. But AMD kept improving.

As of writing this article, AMD controls roughly 40% of server CPU revenue.

AMD’s Switch To TSMC

One of the most important points in AMD’s story is their switch to TSMC starting with Zen 2.

Zen 2 used TSMC 7nm. Zen 3 used refined 7nm. Zen 4 in 2022 used TSMC 5nm. Meanwhile Intel was stuck.

Intel’s 10nm process was delayed for years. It finally shipped in 2019 but had yield problems and lower performance than expected.

Intel’s 7nm was delayed even longer. In 2020 Intel announced another delay and said 7nm would not ship until 2023.

By 2021 Intel was still manufacturing most chips on 14nm. A process they had been using since 2014.

AMD had a process advantage. Zen 3 on TSMC 7nm had better performance per watt than Intel’s chips on 14nm.

AMD’s switch to TSMC has been paramount to their later success. We went deep researching why TSMC is so important for companies like AMD and the rest of the world.

You can read the article that is linked here.

Cloud Providers Join In

And that’s when cloud providers noticed AMD

Amazon AWS launched instances using EPYC in 2018. Microsoft Azure added EPYC options in 2019. Google Cloud followed in 2020.

These companies do not make decisions lightly. They tested EPYC thoroughly. And they found it was cheaper and often faster than Xeon.

By 2021 AMD had 10 percent of the server CPU market. In 2 years that went to 23 percent.

AMD’s data center revenue grew from 1.7 billion in 2020 to 6.5 billion in 2023.
Lisa Su’s bet on chiplets and TSMC had paid off. AMD continued to dominate the consumer side.

Zen 4 launched in 2022 with the Ryzen 7000 series. These chips finally matched Intel in gaming performance while destroying Intel in multi-threaded workloads.

Zen 5 launched in 2024 and maintained the lead.

By 2025 AMD had 30.2 percent of the consumer x86 CPU market. Up from 17.7 percent in 2015.

AMD was making money again. In 2023 AMD’s revenue was 22.7 billion with net income of 854 million.

Their stock hit 180 dollars per share in early 2024. Up from 1.61 dollars in 2015.

But there was still one major problem. GPUs.

AMD vs Nvidia

While AMD was climbing back on CPUs their GPU business was still struggling.

AMD’s market share in discrete graphics cards was 19 percent in 2019. Nvidia had 81 percent.

By 2024 AMD’s share of GPUs had dropped to 12 percent while Nvidia had 88 percent.

AMD’s problem was not raw performance. AMD’s Radeon RX 6000 series in 2020 was competitive with Nvidia’s RTX 3000 series in raw performance.

In some games AMD was faster. In others Nvidia was faster. Overall they were similar. But Nvidia had an ecosystem advantage.

First, Nvidia had better software. Their GeForce drivers were more stable. Game developers optimized for Nvidia cards because that is what most gamers owned.

Second, Nvidia had features AMD lacked.

Nvidia launched ray tracing hardware with RTX 2000 in 2018. AMD did not have hardware ray tracing until RDNA 2 in 2020.

By the time AMD added ray tracing, Nvidia was already on their third generation of ray tracing cores with better performance.

Third, Nvidia had DLSS.

DLSS or Deep Learning Super Sampling launched in 2019. It used AI to upscale games from lower resolutions.

This lets gamers run games at 1080p internally but display them at 4K. Frame rates were much higher with minimal quality loss.

DLSS 2.0 in 2020 was genuinely impressive. It often looked better than native resolution because the AI model could reconstruct detail that simple upscaling would miss.

AMD’s response was FSR or FidelityFX Super Resolution. It launched in June 2021.

FSR was open source which was great. It worked on AMD, Nvidia, and Intel GPUs. Game developers could add it easily.

But FSR 1.0 was not AI-based. The result was that FSR 1.0 looked noticeably worse than DLSS especially in motion.

FSR 2.0 in 2022 looked much better but still lagged DLSS 2 in image quality.

AMD was always one or two years behind Nvidia. By the time AMD matched a Nvidia feature Nvidia had moved on to the next thing.

Fourth, Nvidia dominated AI.

The AI boom that started in 2022 made Nvidia’s GPUs essential for training large language models.

Every AI lab wanted Nvidia H100 GPUs. These chips cost 30000 to 40000 dollars each and were in such high demand that delivery times stretched to months.

Nvidia’s data center revenue was 18.4 billion in fiscal 2023. It hit 47.5 billion in fiscal 2024. And it exploded to over 100 billion in fiscal 2025.

AMD had GPUs for AI. On paper they were competitive with Nvidia but nobody wanted them.

Because the problem was software. Every AI framework was built on CUDA. PyTorch, TensorFlow, JAX all used CUDA.

Researchers had years of CUDA code. Universities taught CUDA. Every AI tutorial assumed Nvidia GPUs.

AMD’s ROCm was supposed to be their CUDA equivalent. But ROCm was buggy and poorly documented for years.

By the time AMD fixed ROCm the ecosystem had moved on. Nobody wanted to port their code from CUDA to ROCm.

Some companies bought AMD MI300 chips to hedge against Nvidia supply shortages. But given the choice everyone picked Nvidia.

AMD’s data center GPU revenue was around 400 million in 2023. Nvidia’s was 47.5 billion.

On gaming GPUs the situation was similar. AMD had 12 percent market share. Nvidia had 88 percent.

AMD was getting crushed still but last year something has changed.

Is AMD Finally Catching Up to Nvidia?

Last year AMD launched the latest version of their upscaling technology - FSR 4.

FSR 4 was completely different from FSR 3. AMD rebuilt it from scratch using machine learning just like DLSS.

The FSR 4 model was trained on thousands of game frames at multiple resolutions. It learned how to reconstruct high resolution detail from low resolution input.

Reviews of FSR 4 were positive. Digital Foundry said it was “the closest AMD has come to matching DLSS.”

Tom’s Hardware called it “finally competitive.”

In side-by-side comparisons FSR 4 on RDNA 4 was very close to DLSS 3.5. Not quite as good but close enough that most people would not notice the difference.

For the first time since 2019 they had an upscaling solution that genuinely competed with Nvidia.

In March 2025 AMD launched RDNA 4 with the RX 9070 XT. This GPU was positioned as a midrange card at 599 dollars. And it competed with Nvidia’s RTX 5070 Ti at 749 dollars. (Though actual end prices may differ due to market conditions.)

Raw performance was roughly equal and ray tracing gap was smaller than previous generations.

With FSR 4 enabled the RX 9070 XT often matched or beat the RTX 5070 Ti in overall gaming performance.

On the data center side AMD launched its MI355X accelerators in mid-2025.

Some big customers are finally willing to try AMD for AI.

Meta announced in January 2026 that they bought thousands of these chips for Llama 4 training. Microsoft said they were deploying MI chips in Azure.

It is not enough to challenge Nvidia’s dominance. Nvidia still has over 90 percent of the AI accelerator market.

But AMD’s data center GPU revenue hit 3.5 billion in fiscal 2025. Up from 400 million in 2023.

Progress is slow on this front but it is real.

Gaming Consoles Keep AMD Alive

There is one part of AMD’s business that does not get headlines but matters a lot for the company and that is console chips.

• PlayStation 5 and Pro use a custom AMD chip.

• Every Xbox Series X and Series S uses an AMD chip.

• The Steam Deck uses an AMD chip.

• Valve’s upcoming Steam Machine uses an AMD chip.

Sony and Microsoft chose AMD because AMD was the only company that could provide competitive CPU and GPU in a single package.

Intel had great CPUs but their integrated graphics were mediocre. Nvidia had great GPUs but no x86 CPU license.

Only AMD could do both.

The console business does not have great profit margins. Sony and Microsoft negotiate hard on price because they buy millions of chips.

But it provides stable revenue.

AMD’s gaming segment which includes consoles and semi-custom chips generated 6.2 billion in revenue in fiscal 2023. That was 27 percent of AMD’s total revenue.

More importantly consoles kept AMD’s GPU architecture relevant during years when they were losing badly to Nvidia on PC.

Game developers had to optimize for AMD GPUs because the PS5 and Xbox use AMD graphics. Those optimizations helped AMD’s PC GPUs run games better.

When Valve launched the Steam Deck in February 2022 they chose AMD for the same reason. AMD could provide CPU and GPU in a compact power-efficient package.

The Steam Deck was a huge hit. Valve sold millions of units. It proved handheld PC gaming was viable.

Now Valve is working on the Steam Machine. This is essentially a console-like PC for the living room running SteamOS.

It uses an AMD chip. If it succeeds AMD gets another high-volume customer.

Consoles will never be a huge profit driver. But they provide a floor. A baseline revenue that keeps AMD stable when CPUs or GPUs are struggling.

For a company that has nearly gone bankrupt multiple times, stability is invaluable.

Can AMD Actually Win or Just Survive

At the time of writing this, here is how things are for AMD.

On CPUs they are winning. AMD has roughly 30 percent consumer market share and 24 percent server market share as of 2025.

AMD’s EPYC processors are taking consistent share in data centers.

Intel is still bigger. Intel’s 2025 revenue was around 50 billion. AMD’s was around 26 billion.

Intel’s manufacturing is still recovering from years of delays. Their 18A process launched in early 2026 but yields are still ramping.

AMD is using TSMC’s 3nm and soon 2nm. Those processes are proven. AMD does not have to worry about manufacturing.

On GPUs AMD is losing but improving too. They have 14 percent of the gaming GPU market. Nvidia has 86 percent.

FSR 4 closed the gap with DLSS. RDNA 4 is competitive in the midrange. But Nvidia’s high-end GPUs are unmatched.

In AI accelerators, Nvidia is the only king with over 90 percent market share. AMD is not going to beat Nvidia in AI anytime soon. Maybe not ever.

But AMD does not need to beat Nvidia to survive. They just need to capture 5 to 10 percent of the market.

If AI spending hits 300 billion per year by 2028 and AMD captures 8 percent that is 24 billion in revenue. That would be transformative for AMD.

AMD has survived for 57 years. They have outlasted countless competitors. Cyrix. Transmeta. Via. All gone.

AMD came back from the brink in 2015. They could do it again if needed. But surviving is not the same as winning.

AMD’s goal is not to be the biggest. It is to be good enough to stay in the game.

Lisa Su took a company worth 1.5 billion dollars and turned it into a company worth close to a trillion dollars in 14 years.

She killed the losing mobile business. She bet on Zen when everyone thought AMD was finished. She committed to chiplets when everyone said monolithic was the only way.

She partnered with TSMC when AMD’s own fabs were failing. She focused on servers when everyone said AMD could never compete with Intel in data centers.

AMD could have been bankrupt a long time ago but it is still here.

The Underdog That Must Fight

As someone who’s writing this article on an AMD GPU, I want the company to exist.

Look, I see AMD as an underdog that fights well above its weight. I mean this small company is fighting the two behemoths in the CPU and GPU domains, all at the same time.

AMD has to exist, not because it may just overtake Nvidia but because it keeps Intel and Nvidia on their toes.

These big companies don’t have much incentive to compete when the underdog disappears.

Ryzen is among the biggest reasons why Intel is redeveloping its manufacturing process and designs because they know they’ve lost to Ryzen.

AMD doesn’t have the best technology in all its domains but it doesn’t have to have that. If it keeps up a good fight, it has done its job.

AMD learned their lessons the hard way.

They lost focus in the 2000s trying to do too many things. Mobile chips. Low-end graphics. Bulldozer’s flawed design.

Under Lisa Su AMD became focused. Three markets: gaming, data centers, embedded and that focus is what has saved them.

The next 10 years will test whether AMD can move from survival to actual victory.

Can they take 40 percent of the CPU market? Can they get to 25 percent of GPUs? Can they grab 10 percent of AI?

Nobody knows. Maybe not even AMD but I sincerely hope they keep fighting.

As always, your feedback is what improves our writing here at SK NEXUS. We’re a small publication funded by no one but ourselves.

Please feel free to share the publication to that curious mind who would find it useful.

Share SK NEXUS

Similar Articles from SK NEXUS

Learn More

• The rise and fall of AMD: How an underdog stuck it to Intel

• The rise and fall of AMD: A company on the ropes

• Bulldozer architecture analysis

• AMD Settles FX Bulldozer False Advertising Lawsuit for Roughly $35 a Chip

• From the brink of bankruptcy to a 1,300% stock gain

• Lisa Su Saved AMD. Now She Wants Nvidia’s AI Crown

• What AMD Learned From Its Big Chiplet Push

• AMD’s Next GPU Is a 3D-Integrated Superchip

• AMD reaches 46% of server x86 CPU revenue

• AMD FSR 4 vs. Nvidia DLSS 4: Shockingly close

• DLSS 4.5 vs FSR 4 – Which is Better for Image Quality?

• FSR 4 vs DLSS 4: It’s closer than you think

• AMD’s RDNA4 GPU Architecture at Hot Chips 2025

• Microsoft Partners with AMD for Next-gen Xbox Hardware

• 3 reasons gaming handhelds favor AMD chipsets

• Could AMD break Nvidia’s chokehold on chips?

• How AMD Left GlobalFoundries for TSMC

Some time ago, I covered prediction markets, what they are, how they work, and why they’ve suddenly become part of mainstream conversation.

At the time, I thought that was enough. Another weird internet-native financial product sitting somewhere between gambling and politics. But then I ended up watching one of Nilay Patel’s podcast discussions around prediction markets, and it pushed me back into the rabbit hole again.

And to me, this whole thing feels strangely familiar.

Think about where we were around ten years ago. Sports betting ads were controversial. Gambling companies weren’t deeply embedded into mainstream entertainment yet. Today, betting platforms sponsor broadcasts, creators openly promote gambling apps, and odds are treated like part of the viewing experience itself.

Now look at prediction markets.

We’ve reached a point where market odds are being cited as news before the event itself has even happened. A politician makes a statement, odds move instantly, and suddenly the movement becomes part of the story.

And that’s really the purpose of this piece, to look at what happens when speculation starts becoming the lens through which people process reality itself.

The Pseudo-Event Problem

Traditionally, news had a fairly simple foundation - something happened, reporters verified it, evidence emerged, and coverage followed. The sequence mattered because journalism was tied to documenting events that had already occurred.

But modern media ecosystems introduced another category entirely - Pseudo-events.

These are events designed primarily to generate attention and coverage rather than communicate something meaningful. Press conferences teasing announcements, staged debates, “major updates coming soon,” carefully timed leaks…etc.

So what’s the real purpose of these pseudo-events? It’s to create anticipation & engagement around the possibility of something happening.

Prediction markets fit perfectly into this environment because they thrive on uncertainty and the markets just need people to guess on whether it might happen or not.

You can already see how the cycle forms.

A political campaign announces that a “major policy change” is coming next week. Prediction market odds react immediately. News outlets begin reporting that movement as if it reflects meaningful new information. Social media discussions explode around the odds change itself.

Then the announcement gets delayed or watered down, but by that point, it doesn’t really matter because what’s done is done.

And that changes the role of journalism in subtle ways.

Instead of reporting on verified developments, coverage begins orbiting around speculative momentum. The narrative becomes the event. Odds movement becomes the update and overall, the market sentiment becomes something reporters feel pressured to include because everyone else is watching it too.

How Prediction Markets are Reframing Themselves as Truth Infrastructure

The language around prediction markets has changed a lot over the last few years.

Instead of “gambling,” the industry talks about “forecasting,” “market signals,” and “information aggregation.”

Polymarket itself describes itself as a system that uses markets to make sense of real-world events. Kalshi uses the language of “event contracts” and regulated forecasting markets, emphasizing legality and data rather than speculation.

And quite frankly, the reframing works because if you’re a news outlet trying to support a story with numbers, prediction markets will do the job, just fine because the odds look pretty similar to polling data. On top of that, both move over time and both appear to reflect public expectations.

But they measure completely different things.

Polls measure what people say they believe. Prediction markets measure where money is flowing based on what participants think might happen and this distinction matters a lot more than people realize.

During the 2024 U.S. presidential election cycle, prediction market odds started appearing in mainstream political coverage alongside traditional polling. Headlines discussing changes in Polymarket odds were presented almost like electoral indicators, which these markets are not.

And this creates a loop very pretty quickly - A market moves. News outlets cover the movement. Readers see that coverage and assume something important changed. More people enter the market, the odds update again, and another round of coverage follows.

Before you notice, the conversation revolves around market activity itself.

This is also where pseudo-events fit naturally into the system. A campaign hinting at a “major announcement,” can move odds within minutes, even if nothing concrete actually exists behind it.

The coverage still happens and it becomes clear that the whole world is reacting to narratives moving through the media cycle. And when those reactions start getting treated as cues for something important, it becomes really hard to separate actual reporting from speculations.

Digital Déjà Vu

A lot of this feels familiar because we’ve already seen a very similar transition happen with sports betting.

The language around sports betting slowly evolved over time.

Betting platforms stopped presenting themselves purely as gambling products and started positioning themselves as part of the viewing experience. Odds became live analytics. Gambling became fan engagement and real-time betting became interactive sports entertainment.

Today it’s difficult to watch major sports broadcasts without betting odds appearing directly on-screen. These companies sponsor teams, creators, podcasts, and even entire segments during live coverage. The integration became so normalized that for younger audiences, it barely even registers as gambling anymore.

Now look at how prediction markets are being framed.

The same vocabulary transition is happening again and the underlying behavior hasn’t changed much. People are still putting money behind uncertain outcomes.

Prediction markets operate much closer to real-world events. Elections, regulations, wars, economic policy, company decisions - all of it becomes something users can financially bet on.

That changes how these systems interact with public perception. These markets are in real-time influencing how people process information itself due to which a market movement is soon going to look like some kind of signal about reality.

And that creates a very different kind of environment compared to traditional gambling platforms.

The Gacha Parallel

This pattern also resembles what happened with Gacha systems and loot boxes in gaming.

For years, games introduced mechanics built around randomized rewards, especially in mobile titles. Players spent money for a chance at obtaining specific characters or items, often through systems designed around probability and repetition.

A large part of the audience exposed to these systems was underage.

Regulatory responses came slowly, and by the time serious discussions started happening around loot boxes and gambling mechanics, millions of players had already grown up with these systems as a normal part of digital entertainment.

I covered this in much more detail in my gambling article, so I won’t go too deep into it again here.

But prediction markets follow a surprisingly similar structure.

Betting on things has been gamified. Platforms encourage users to track outcomes, compare positions, share predictions, and continuously engage with unfolding events. And unlike Gacha systems, prediction markets aren’t confined to gaming communities because these markets target anyone consuming information.

That’s the scale difference - sports betting monetizes attention around sports. Gacha systems monetize progression and reward loops inside games. Prediction markets monetize uncertainty surrounding real-world events themselves.

The Future of Market-Driven Information

The bigger concern here is that - what happens when prediction markets become embedded into everything else.

We’re already seeing the early stages of it. Prediction market widgets appear inside social feeds. Creators reference market odds during discussions. News outlets cite betting probabilities alongside reporting.

Furthermore, political campaigns are beginning to optimize for market reactions. Headlines are written around odds movement because it generates engagement. Financial incentives are starting to shape public discourse at an individual scale.

For younger audiences growing up inside these systems, checking the odds is going to become more of a common reflex than verifying the source.

And this ties directly into the ecosystem discussion we had earlier.

Prediction markets are following the same pattern most modern platforms follow. First comes integration, after that normalization & dependency follow. And the more deeply odds become embedded into feeds and conversations, the harder it will become to separate speculation from the way people interpret events altogether.

Why This is Worse Than You Think

Fake News Becomes Profitable at Individual Scale

Right now, misinformation mostly runs on attention. People spread false or misleading content because it generates clicks & engagement.

Prediction markets introduce another layer on top of that which adds a direct financial incentive attached to shaping public sentiment itself.

In November 2025, a staffer at the Institute for the Study of War made an unauthorized edit to one of the organization’s maps showing Russian forces capturing a key intersection in the Ukrainian town of Myrnohrad. That was a big move because Polymarket relied on those maps to settle war-related contracts.

Because of that, users who had bet on Russia taking the city reportedly made returns as high as 33,000%, with the total market reaching around $1.3 million. Later, the edit disappeared. The Institute for the Study of War (ISW) stated that the change had not been approved and removed it from the next day’s workflow. Shortly after, one of the organization’s geospatial researchers disappeared from the website and reports indicated the staffer had been fired.

The important part here is the realization that if somebody can manipulate information connected to a market, it becomes possible to financially benefit from influencing real-world situations tied to those bets.

Markets Can Be Gamed by Coordinated Actors

Another issue is how easily markets can create artificial signals.

A small but coordinated group placing concentrated bets can move odds significantly, especially in thinner markets. Once those odds move, social media notices. News coverage follows. People outside the market start interpreting the movement as meaningful information.

And the cycle feeds itself from there.

We already saw a version of this during the Maduro case covered in Part 1.

In early 2026, a Polymarket account placed large bets predicting Venezuelan leader Nicolás Maduro would lose power before the end of the month. The timing immediately raised eyebrows because the trades were placed only hours before a surprise U.S. military operation ultimately removed him from office.

When the outcome was confirmed, the account reportedly walked away with close to $400,000 in profit.

What made the situation stranger was the account itself.

It had very little trading history and appeared to exist almost entirely around those specific wagers. Naturally, people started asking whether the trader simply got extremely lucky or whether the bets reflected access to information the public didn’t have.

And even if no direct wrongdoing occurred, the bigger issue remains the same.

Prediction markets create environments where market activity itself starts behaving like a signal. A sudden odds movement can influence public perception long before anyone knows whether the underlying information is reliable which can create a system where coordinated actors just need enough momentum to make people believe something important is happening.

What You Should Keep In Mind

Understanding What You’re Looking At

The first thing to understand is that prediction market odds are not truth indicators. They just reflect where money is moving. So when you see headlines citing market odds as evidence of something becoming “more likely,” it’s worth asking a few questions:

Who benefits if public perception changes?
Who is placing the bets?
And what incentives exist behind the movement itself?

Markets can aggregate sentiment, but sentiment and reality are not the same thing.

In short, a percentage attached to an outcome does not automatically make the underlying claim more credible which is why verification still matters

Don’t Let Odds Replace Sources

Prediction markets become much more influential when they get embedded directly into the platforms people already use.

Odds inside social feeds, prediction widgets attached to news coverage, creators casually referencing betting percentages during discussions - all of this gradually conditions audiences to treat markets like informational shortcuts which plays a key role in forming a new habit that involves people depending on odds instead of credible sources.

So what’s the solution?

If platforms give you the option to disable prediction widgets or betting integrations, turn them off. If creators constantly frame discussions around market activity, recognize with your whole heart that you’re being pushed toward interpreting speculation as meaningful information.

Closing Thoughts

Prediction markets are moving through a very similar process to what we have seen in the past with the whole gambling matter.

If these systems become normalized, reversing it would be much more difficult.

And all of this is happening in an internet environment that is already messy enough.

We’re surrounded by AI-generated content, bots, misinformation, deepfakes, algorithmically amplified narratives - distinguishing between reality and manipulation is already getting harder by the day.

Prediction markets add another layer to that environment and the substructure around it is still being built.

That’s why awareness matters now, before these systems become deeply embedded into the way people consume information every day.

Anyway, thanks for reading.
See you next time.

Most people use technology every day without ever stopping to ask how it works, who controls it, or what happens when things go wrong.

This publication exists for the opposite reason.

So, if you enjoy deep dives into emerging technology, privacy, AI, cybersecurity, and the systems shaping the modern world, consider subscribing.

That way, the next article will show up directly in your inbox.

What You Might Like

• Ensh*ttification - How Platforms Turn Against You

• What The Heck is AI

• The Business of Trust - Influencer Marketing

• The Death of Ownership: Why Your Tech Isn’t Really Yours Anymore

It’s good to have you back.

While working on my last piece on digitization of gambling , I kept running into something strange during research - Prediction markets.

At first, I didn’t think much of it. Then I started seeing odds for elections, wars, policy decisions, even oddly specific future events. And later I realized that these were entire markets.

The deeper I looked, the more unavoidable this became. Instagram reels, YouTube shorts. Clips of people bragging about how much they made by betting on real-world outcomes. Someone won because they predicted this election. Someone else cashed out because they bet on that crisis.

A lot of questions started to pile up in my head.

Why are people trusting markets to predict reality? Why do odds feel more convincing than experts or institutions? And when did forecasting quietly turn into a form of gambling wrapped in data and probability?

And once again I guess, I decided to find the answers to those questions.

You see…gambling doesn’t always look like a casino. Sometimes it looks like insight, sometimes it looks like intelligence. And sometimes, it slips into daily life without you noticing.

In this article, I’ll break down what prediction markets actually are, where they came from, what their growing popularity says about how we understand the future and much more…

Every second counts, let’s continue.

What Prediction Market Actually is

Polymarket or Kalshi are prediction markets. In other words - they are gambling platforms dressed up as forecasting.

On the markets - users are able to bet real money using the USDC stablecoin on who would win. You can wager on everything from the next James Bond actor to whether the U.S. government will confirm the existence of aliens.

What sets Polymarket apart is perception. It’s now regularly cited by mainstream media and amplified by high-profile figures like Elon Musk, who has argued that polymarkets beat polls because real money forces honesty.

At its core, it works by letting people buy and sell shares tied to future outcomes. If the event happens, the share pays out. If it doesn’t, it goes to zero. The price reflects what participants collectively believe, not what’s guaranteed to happen.

Where Prediction Markets Came From

Humans have always tried to price the future. Long before spreadsheets, people were informally betting on wars, royal successions, political outcomes...etc In the 1500s and 1600s, political betting markets were already operating in coffeehouses across Europe.

But it wasn’t as smooth as it is right now.

Prediction markets as a serious concept didn’t emerge until the late 20th century.

In 1988, economists began formalizing the idea that markets could aggregate scattered information better than polls or expert panels. This thinking led to the launch of the Iowa Electronic Markets in 1989, a small academic experiment that allowed participants to trade contracts tied to election outcomes.

Throughout the 1990s and early 2000s, prediction markets gained academic validation. Economists studied them as forecasting tools, not gambling systems. Corporations experimented with internal markets to predict product launches, sales targets, project delays…etc

The idea behind the prediction market was clear - when people have skin in the game, they reveal information more honestly which most of the times lead to more accurate results.

So yes…prediction markets were initially built for forecasting and over time, everything got flipped.

The real public boom came decades later. During the 2024 U.S. presidential election, prediction markets were suddenly everywhere. Odds were screenshotted, shared on social media, cited by journalists, and compared directly with polls. The election acted like a spotlight for prediction markets and that moment changed everything…

How Prediction Markets Gained Momentum

The main reason behind why prediction markets exploded is - the change of environment.

Back in the ol’ days, there was no internet.
But that isn’t the case now.

The internet removed the old gatekeepers. You no longer needed academic credentials, or any permission to participate. At the same time, regulation is slower than ever. Prediction markets didn’t fit neatly into existing gambling, which gave platforms room to operate in legal gray zones while regulators played catch-up.

Crypto accelerated everything. Stablecoins made global participation, bypassing banks, borders, and payment processors frictionless. Suddenly, anyone with an internet connection could participate in markets tied to real-world events.

Once a market starts gaining traction, social media does the rest. Screenshots spread. Wins go viral. Algorithms amplify what performs well, nudging more people towards what it wants them to see.

On top of that - Journalists notice what’s already trending, they along with influencers also start to hop in the shiny train ultimately giving birth to a never ending loop.

That’s exactly what happened with prediction markets.

And if you want to understand how platforms quietly shape what you see, what you believe, and what you pay attention to, I broke it down in detail here:

How Prediction Markets Work

At a high level, prediction markets turn beliefs into prices.

Each possible outcome is represented by a share. When more people believe an outcome will happen, demand for that share increases and its price rises. When confidence drops, the price falls.

As new information appears, let’s say a court ruling or breaking headline or a sudden injury, prices adjust almost instantly. People who react early and correctly are rewarded. Those who guess wrong lose money. That incentive structure is the entire system.

So basically, you’re buying into a probability shaped by the crowd. If the outcome happens, the share pays out. If it doesn’t, then you lose money.

What you’re really seeing is a live signal of collective belief, constantly updated as people put real stakes behind their assumptions. And the market only cares whether you are right or not, it doesn’t give two cents about why you believe something.

Polymarket is also called ‘wisdom of the crowds’ though, it’s partly true because the information can be manipulated via insider trading which is the biggest problem of this system about which we will be talking next.

What the Problem with Prediction Market is

The real problem with prediction markets is the asymmetry.

Imagine a streaming platform sitting on a yet-to-be-released finale. There’s a market betting on whether a major character dies. Writers, editors, interns - dozens of people already know the answer. Any one of them can quietly place a bet days or weeks in advance and walk away with easy money.

That’s literally information leakage with a payout.

Some markets are genuinely uncertain. Elections, championships…etc no one has the final answer beforehand. But many Polymarket questions aren’t like that. They revolve around private relationships, unreleased announcements, medical events, or corporate decisions. In those cases, someone always knows what’s going to happen next.

Markets asking whether a celebrity couple will split, or if some celebrity is going to get pregnant or whether a famous CEO will step down…etc These are invitations for insiders, friends, assistants, or employees to cash in on knowledge the public doesn’t have.

This turns many prediction markets into sucker’s games. You think you’re participating in a neutral forecasting tool. But you’re just trading against people who already have the answer.

The Grey Zone: Forecasting or Just Gambling?

Prediction markets are sold as sophisticated forecasting mechanisms. The language is polished to make it sound like Wall Street rather than DraftKings - “probability signals”, “information efficiency”, “decentralized knowledge”, “wisdom of crowd”...etc.

Yes, the mechanics are dressed in financial language. Yes, there are probability charts instead of point spreads. But under the hood, it’s simple - risk capital on uncertainty, win if correct, lose if not.

Regulation reflects this ambiguity. In some regions, prediction markets are treated like financial derivatives. In others, they’re classified as gambling. In many places, they operate in a gray zone where neither framework cleanly applies. This inconsistency is a clear reflection of what these markets actually are.

And if you’ve been paying attention to how gambling has been digitalized over the last decade, this should feel familiar.

Fun fact - I recently broke down how gambling is moving online, embedding itself inside video games and different consumer apps. If you want the deeper breakdown of how digital gambling is evolving and realize how prediction market is just another form of gambling, you can check this post out:

How Prediction Markets Are Blending Into Content

Prediction markets are beginning to intersect with the creator economy in subtle ways. These markets aren’t any niche financial tool anymore, discussions about them are increasingly appearing in livestreams and social media.

Streamers do not need to actively place bets to influence participation. Simply discussing probabilities or live market movements is more than enough for promoting these markets.

For example - Atrioc, a youtuber, whose audience already gravitates toward economics and probability-driven analysis, represents the kind of community where prediction markets naturally gain traction - even if markets are not the primary focus of the stream.

The overlap is expanding beyond commentary. Sports figures are also entering the space. Giannis Antetokounmpo recently joined prediction platform Kalshi. This is proof of how the market is slowly growing.

Meanwhile, platforms such as Polymarket routinely host markets tied directly to internet personalities.

At this point, it’s important to ask - whether viewers are consuming commentary or are they interacting with tradeable outcomes linked to personalities they follow? Makes sense right?

Comment down below and let me know what you think

Leave a comment

Just to be clear - none of this is inherently ‘wrong’. But the blending of entertainment, and monetized markets raises practical questions. If a streamer discusses a market, is it organic curiosity or sponsored promotion? In a digital ecosystem where affiliate relationships are not always obvious, audiences may struggle to distinguish between commentary and marketing.

This issue is a whole thing in itself - I explored this dynamic in more detail in a previous piece on how online trust itself has become monetizable and why not every recommendation is neutral:

Case Studies That Made People Question Things

Let’s take a look at real cases that grabbed attention and forced people to think about what goes on behind the scenes of this market.

The $400,000 Maduro Contract That Drew Attention

In early January 2026, one of the most striking stories to emerge from the prediction market world centered on a massive payout tied to Venezuelan leader Nicolás Maduro’s sudden removal from power.

A Polymarket account placed significant wagers that Maduro’s leadership would end by month’s end and those trades were made just hours before a surprise U.S. military operation ultimately removed him from office. When the outcome was confirmed, the bets paid off at near-full value, generating a profit of nearly $400,000.

What made the story so crazy was the timing and the mystery surrounding the bettor. The account had been created only weeks earlier and had little trading history outside of these specific wagers. That led many online observers and analysts to ask the same question:

“Was this luck, or did someone have access to privileged information that the broader public and markets didn’t?”

NFL Banning Prediction Market Ads During the Super Bowl

Prediction markets have officially been placed on the NFL’s list of prohibited advertising categories, according to Front Office Sports. A source familiar with the matter told the outlet that these ads were barred not just from the Super Bowl, but from NFL broadcasts throughout the 2025 season.

The Super Bowl is the most-watched television event in the United States. Last year’s broadcast drew an average of 127.7 million viewers across platforms - the largest audience ever recorded for a single U.S. television event, according to Front Office Sports. Thirty-second ad spots have climbed as high as $10 million this year.

So why block prediction markets?

The answer is pretty simple.

The NFL has spent years carefully managing its relationship with gambling. Sports betting is now integrated into broadcasts, but it’s tightly regulated and politically normalized. Prediction markets operate in a much shadier zone.

They don’t sit cleanly inside gambling law.
They don’t sit cleanly inside financial regulation either.

This ambiguity creates legal and reputational risk and the NFL does not gamble with reputational risk during its events.

Allowing prediction market ads during the Super Bowl would effectively place them alongside established sportsbooks. That’s a stamp of approval - A signal to the public that these platforms belong in the same category.

This decision is pretty subtle, but it reveals something bigger if you look closer - Major institutions are cautious about prediction markets because they understand them very well. This matter suggests that even powerful institutions see them as closer to gambling than forecasting and them taking this decision is a sign of that belief.

What This Means Going Forward

I believe prediction markets deserve a more balanced assessment.

Platforms like Polymarket have proven to be useful as information tools. When people put their own money behind an outcome, that creates a stronger signal than a traditional poll. As Elon Musk has pointed out, markets can sometimes be more informative than surveys because participants have financial exposure. Incentives matter.

However, a priced probability is still just a probability.

At the same time, their expansion is undeniable. As reported by Forbes, trading volume across major prediction platforms surpassed $3 billion in Q3 2025 alone - more than five times higher than the same quarter a year earlier. Industry estimates project the sector could reach nearly $95.5 billion by 2035, with annual growth rates approaching 47 percent.

And now, the very platform you’re reading this on - Substack - has partnered with Polymarket. Writers can embed live prediction widgets in Notes, showing odds for elections, policy moves, or even celebrity drama, all with real money on the line.

This level of growth suggests prediction markets are growing into something bigger.

In my view - prediction markets will continue to expand, and their influence will likely grow. But growth does not resolve their structural ambiguity between gambling and finance. And expansion does not eliminate uncertainty.

What remains unsettled is how regulators, institutions, and the public will ultimately choose to classify prediction markets. That decision may determine whether prediction markets become embedded infrastructure or not…

Because once prediction markets start influencing journalism and the way people interpret information itself, the conversation changes completely.

And that’s exactly what we’re going to explore in Part 2, stay tuned for that!

One Last Thing

Nearly all people interact with systems without questioning them. This space is about doing the opposite aka understanding how things work, where they fail, and what that means in practice.

If that’s the kind of content you want more of, subscribe. And if this piece added something new to how you see things, pass it on. Good information only matters if it travels.

And if you’re active elsewhere, follow along on Bluesky or X (Twitter) or LinkedIn - those are the two places I show up most consistently outside Substack. A follow, a like, a repost - it matters to me, more than you think.

What You Might Like

A while ago, I wrote about hacking as a lifestyle, breaking down what hacking actually is and how it’s more about mindset than just code. But there’s a bigger misconception that keeps showing up.

Nearly all people think hacking started when the internet came into being, but that’s not true.

Hacking started with a whistle and a phone line. It started long before modern cybersecurity, the first hackers were experimenting with telephone systems and were trying to figure out how a global communication network actually worked.

So how did that even happen?
How did a system built for communication become something people could manipulate? And why does it still matter today?

That’s what this piece is about.

But before we get into the history, you need to understand one term.

What Phreaking Actually Is

Phreaking is one of the earliest forms of hacking. It sits at the intersection of curiosity and exploitation, focused entirely on telephone networks.

The term itself comes from combining phone and freaking, which gives you a sense of what it represents - people experimenting with systems they weren’t supposed to touch.

At its core, phreaking was about understanding how telephone systems worked and then using that knowledge to manipulate them. Sometimes that meant making free long-distance calls. Sometimes it meant accessing parts of the network that were never meant to be public.

What made this possible was how early phone systems were designed. Those phones weren’t like the ones we have now, they relied on tones.

Every action in the network such as routing a call, switching lines, connecting across regions, was controlled by specific audio signals. Back then the system didn’t verify identity and blindly trusted the signal behind it.

Phreakers figured this out and reverse engineered those tones and started reproducing them. Once you could generate the right frequency, you could tell the network what to do.

The Origins of Telephone Hacking

How Phone Systems Worked

Long before computers were everywhere, the global communication system ran on sound.

Early telephone networks relied on what’s called in-band signaling. This meant the same line that carried your voice also carried the control signals that told the network what to do. Things like: routing a call, marking a line as busy, connecting long-distance, all of these, depended on specific tones moving through the line. These telephone networks had a major problem, they assumed all those tones being received by it were always legitimate.

Somewhere in the late 1950s and early 1960s, a few curious individuals started noticing the flaw in the system and wondered if the system responded to tones, what would happen if you reproduced them yourself?

That question was the starting point.

The First Phreaker: Joe Engressia (Joybubbles)

One of the earliest figures in this space was Joe Engressia, later known as Joybubbles.

He was born blind and early on he developed an unusually sharp sense of hearing. As a child in the late 1950s and early 1960s, he began experimenting with telephone tones because he wanted to understand them more and due to these experiments, he discovered something critical.

He was able to whistle at a frequency of around 2600 Hz, a tone used by the telephone system to signal that a line was idle. When he produced this tone into the phone, the network responded. Calls could be disrupted, redirected, or reset.

This proved that the system could be controlled using sound.

His work didn’t stay isolated. It spread through early communities and inspired others who were trying to understand the same system. According to early telecom histories and accounts documented in 2600 Magazine, Engressia’s discoveries became a foundation for later phreaking techniques.

John Draper (Captain Crunch) Taking It Public

Put it this way, if Engressia proved it was possible, John Draper made it widespread.

In the early 1970s, Draper discovered that a toy whistle found in Cap’n Crunch cereal boxes could produce the same 2600 Hz tone. The whistles were originally used by sailing officials to signal mealtimes or command, but somehow - one of them fell into the hands of John Draper, a former U.S. Air Force electronics technician.

The uniqueness of this whistle involved producing certain tones that allowed the whistler to bypass AT&T’s analog system and get free long-distance phone calls.

Draper’s work helped shape an early underground culture around telephone hacking. His influence extended beyond telecom systems. Figures like Steve Wozniak and Steve Jobs famously built and experimented with blue boxes before founding Apple, a story later referenced in multiple tech retrospectives.

The Core Vulnerability of the System

The reason all of this worked comes down to design.

Telephone networks used in-band signaling, where control signals traveled along the same path as voice communication. A 2600 Hz tone told the system a line was idle. Additional tones could then direct where the call should go.

This vulnerability existed for decades. It wasn’t fully addressed until the introduction of out-of-band signaling systems like Signaling System No.7 (SS7), which moved control signals off the voice line, but by that point, the damage had already been done.

The Tools That Broke the System

Once people understood that telephone networks responded to tones, the next step was obvious. Phreakers knew that if they wanted consistent control, they needed something more precise.

That’s when phreaking moved from curiosity-driven experiments to something more structured. Instead of trying to imitate tones, hobbyists started building devices that could generate them exactly. Over time the knowledge started spreading and tools began to standardize.

Blue Boxes: The Turning Point

The most important of these tools was the blue box. A blue box was a device designed to reproduce the exact tones used by telephone systems for long-distance routing. It allowed users to send precise signals into the network and direct calls however they wanted.

In practical terms, this meant you could take control of a call after it connected. By sending the right sequence of tones, the system would treat you like an internal operator. Calls could be rerouted, extended, or made across long distances without being billed correctly.

John Draper played a key role in popularizing these devices, but the idea didn’t stop there. In the mid-1970s, even Steve Wozniak and Steve Jobs experimented with building and selling blue boxes before starting Apple Inc. That alone tells you how accessible the concept had become.

Other Tools in the System

The blue box was the most well-known tool, but it was part of a larger set of devices.

Phreakers didn’t stick to one approach. As they learned more about how the network behaved, they built different tools to experiment with different parts of it. Black boxes and red boxes are some examples circulating within the same communities.

The Mechanism Behind It

At their core, all of these tools did the same thing - They generated sound.

Telephone networks used predefined audio frequencies to manage connections. A blue box could produce combinations of these tones, allowing users to simulate internal commands. By controlling the sequence, you could influence how a call moved through the network.

That’s what made these tools so powerful. All one had to do was understand how to speak to the system. And for a brief period, the network couldn’t tell the difference between a legitimate signal and a manufactured one.

Jobs and Wozniak’s Phreaking Phase

By now, you’ve probably noticed that whenever phreaking comes up, the names Steve Jobs and Steve Wozniak aren’t far behind. Their involvement shows up again and again, but it’s worth looking at it directly to understand why.

In the early to mid-1970s, Wozniak came across phreaking through the same circles influenced by John Draper and then built his own version of the blue box and improved it to make it more stable and reliable.

What Wozniak built as a technical challenge, Jobs saw as something that could be sold. Together, they started building and selling blue boxes to students. And don’t get it wrong, it wasn’t organized crime - it was two curious outsiders testing a system and realizing they could push it further.

Jobs later summed it up in a line that still gets quoted:

“If we hadn’t built blue boxes, there would be no Apple.”

That same mindset still carries forward. The mindset that forces people to learn how systems behave, where they fail, and what happens when you take control of them. If you want to go deeper into that way of thinking, I’ve already broken it down in more detail here: Hacking as a Lifestyle: Beyond Computers.

The Consequences of Phreaking

The concept of phreaking didn’t stay contained for long. It started spreading like a virus with the passage of time.

And surprisingly, this pattern shows up everywhere. A small group explores a system to understand how it works. They find weaknesses along the way. Then a second group arrives, interested, only in using those weaknesses.

That’s when things begin to change.

By the 1970s, phreaking had moved beyond isolated experimentation. As techniques spread, telecom companies began facing real consequences. Phone fraud was costing companies tens of millions of dollars annually and the bigger issue wasn’t just the money.

Phreaking exposed that the entire system relied on trust, and that trust could be bypassed. Media coverage also started picking up on it. Stories about “phone hackers” began appearing more frequently, bringing attention to a problem tons of users didn’t even know existed.

At that point, ignoring this problem wasn’t an option anymore and once that realization set in, the pressure to completely redesign the system became unavoidable.

The End of an Era

Legal Action Against Phreakers

By the mid-1970s, phreaking, which started as small-scale experimentation, had grown into something telecom companies could measure. AT&T and other carriers began reporting consistent revenue losses tied to toll fraud. According to some sources, from that period, losses were running into millions of dollars annually as long-distance networks were exploited.

After that, the big guys aka law enforcement got involved. Phreaking moved into the category of criminal activity, prosecuted under fraud and unauthorized access laws. Arrests became more common, and well-known figures like John Draper were arrested for wire fraud charges.

Telecom companies also adapted. They began working directly with law enforcement, setting up monitoring systems and sting operations to track down active phreakers.

The System Had to Change

As I have mentioned previously in the piece - In-band signaling had a built-in flaw. It trusted whatever came through the line, as long as it sounded correct. But the moment outsiders figured out how to reproduce those signals, things started going downhill because there was no patch for this.

So the industry moved toward a redesign. Through the late 1970s and into the 1980s, telecom networks began separating control from communication. Instead of sending instructions over the same line as voice, signaling was moved to a dedicated channel.

This change eventually led to systems like Signaling System No. 7 (SS7), where the network no longer relied on tones from the user side to make decisions. And while it fixed the original weakness…it wasn’t perfect and like most large systems, it came with its own trade-offs.

If you want to understand that side of it without going too deep here, this video does a solid job:

How Phone Hacking Evolved

If you look close enough, you’ll realize that phreaking only evolved. The old systems ran on tones. Today’s systems run on software and networks. But the core idea stayed the same. If you understand how a system works, you can find where it breaks.

Take Voice over Internet protocol (VoIP) systems. Instead of physical phone lines, calls now move over the internet. That makes them flexible, but also easier to manipulate. Attackers can spoof caller IDs, reroute calls, or inject traffic into poorly secured systems.

Then there’s SIM swapping. An attacker convinces a telecom provider to transfer your number to a new SIM card. Once that happens, they control your calls and messages. That means access to OTPs, account resets, and anything tied to your number.

Telecom fraud operates at a broader level. According to the Communications Fraud Control Association, global losses reached $38.95 billion in 2023, showing how large this problem has become.

Lessons from Phreaking

Long before cybersecurity became an industry, a small group of outsiders looked at a global system and refused to treat it as something untouchable. They didn’t have formal training or access. But what they did have was - curiosity and time.

Back then there was no internet to learn from. Knowledge only moved through conversations, small circles, and eventually through publications like 2600: The Hacker Quarterly, a culture formed around understanding systems instead of blindly using them. That mindset still shows up today.

Anyway, if you’ve made it this far, you’ve basically walked through that entire phase. From a system that trusted tones to a world where everything is connected and constantly targeted.

Thanks for reading.
See you next time.

One Last Thing

Most people use technology every day without ever stopping to ask how it works, who controls it, or what happens when things go wrong.

This publication exists for the opposite reason.

So, if you enjoy deep dives into technology, privacy, AI, cybersecurity, and the systems shaping the modern world, consider subscribing.

That way, the next article will show up directly in your inbox.

What You Might Like

• Hacking as a Lifestyle: Beyond Computers

• Social Engineering - When People Become the Exploit

• How We Got Online: The Hidden History of the Internet

In the previous part, we talked about ecosystems and why they exist. To be fair, most of that discussion focused on the good side of things. These systems make life easier because everything connects, everything syncs, and you don’t have to think too much about how things work together.

But like everything else, there’s another side to it.
You only start noticing it when you ask a simple question:

“What happens after you’re inside an ecosystem?”

And the strange part is, there’s no moment where the system suddenly turns against you. Everything continues to work, just like before, but the difference shows up slowly.

Small changes get introduced over time. Each one feels minor, almost insignificant on its own. But as they stack, the system starts behaving differently. And because the change is gradual, the majority of people don’t notice it happening.

It’s like that boiling frog analogy. The temperature doesn’t spike. It rises slowly enough that you don’t react to it.

Anyway, let’s get started.

Lock-In Phase

In Part 1, we broke down how ecosystems form. There’s a stage in that process where everything starts connecting, for example: your data, your habits, and the tools you rely on.

That’s where lock-in happens.

Take a simple example.

If you’ve been using Google products for years, chances are you’re not just using one service. You probably have Gmail for communication, Drive for files, Photos for backups, Chrome for browsing, and maybe Android as your operating system.

None of these decisions feel connected when you make them. Each one solves a small problem, but with the passage of time, they start overlapping.

Your files, emails, photos, documents…etc. everything is tied to the same account. Even the way you use your devices starts adapting to that system.

Now add another layer.

The people around you are often using the same tools. Your family shares photos through the same apps. Your work depends on the same platforms. That creates a network effect. Communication becomes easier inside the system, which makes leaving less appealing.

At that point, switching is still possible, but there’s a lot of friction involved in that process because you’re moving years of data, rebuilding workflows, and adjusting habits that took time to form. Most people won’t be willing to go through that due to which they stay because leaving feels like unnecessary effort.

Control Starts Shifting

As discussed earlier - when the users are settled into a platform, the priorities inside that platform begin to change.

In the early stages, everything is built around attracting users. Features are designed to feel open and easy to use. But once a large user base is established, the pressure to compete in the same way starts to reduce.

A clear example of this is how feeds changed over time.

When Instagram launched, posts were shown in chronological order. You saw content based on who you followed and when it was posted. Over time, that changed. The platform introduced an algorithmic feed that decides what you see and in what order. Even when chronological feeds were brought back later, they weren’t set as default. You still have to switch manually each time.

Alternatively, when YouTube introduced autoplay, it was turned on by default. The next video starts automatically, selected by the system which plays a key role in how people consume content without them actively choosing it.

In another case, TikTok was fined €345m in Europe for default settings that pushed younger users toward more public visibility instead of keeping accounts private.

Individually, these changes might seem small, but together, they show a pattern - platforms start by giving users control. Over time, they begin to guide behavior more actively. Not in a way that breaks the experience, but in a way that gradually reshapes it.

There’s a term for this - dark patterns and I have already covered in detail on what these patterns are, how the algorithm is manipulating you and how to protect yourself from it, check it out here:

Case Studies

Although this isn’t limited to one company and different platforms approach it in different ways, the underlying structure stays the same.

And just a side note - the goal here isn’t to label companies as good or bad. It’s to understand how the system works.

Google

If you want to understand how ecosystem lock-in works in practice, Google is one of the clearest examples.

At the center of it all is a single Google account. That account connects almost everything - Gmail for communication, Drive for storage, Photos for backups, Chrome for browsing, Play Store for apps, and a long list of other services. Each of these tools works fine on its own, but the real value comes from how they connect, and that’s where things start to compound.

As time passes, your entire digital footprint begins to settle inside that system and everything starts living in one place. It’s not something you consciously plan. It just builds gradually as you keep using the same tools.

And when you imagine trying to leave. Suddenly, you’re dealing with years of accumulated data. While tools like Google Takeout allow you to export your data, that’s only one part of the process. You still have to move everything, set up alternatives that don’t behave the same way. That friction is enough to stop most people.

Another layer that pulls users deeper into the system is single sign-on. A large number of third-party apps and services allow you to sign in using your Google account. It’s convenient, but this creates a dependency chain. Your access to multiple services becomes tied to that one account, which means if you leave Google, you’ll have to reconfigure access to everything connected through it.

Then there’s the infrastructure layer - even if you try to step outside the ecosystem, say by installing apps from alternative app stores, many of those apps still rely on Google Play Services in the background. Things like notifications, location tracking, authentication, and payments often depend on it. This means that even outside the visible layer, parts of the system still depend on Google’s infrastructure, which makes full separation harder than it seems.

There’s also a less obvious dependency that builds over time. Google updates Play Services independently of Android system updates. That means even older devices, which may not receive full operating system updates anymore, still rely on Google’s services to function properly.

Storage plays another role in this. Google offers around 15GB of free storage shared across Gmail, Drive, Photos, and other services. At first, that feels like more than enough. But as your usage grows, that limit gets reached. At that point, you’ll either clean up years of stored data or move to a paid plan like Google One. Most people choose the second option, ultimately making leaving even more difficult now the person has to deal with both data migration and subscription dependency.

Then there’s the data layer - Google’s services improve as they collect more user data. From a user perspective, this often feels like improvement. From a system perspective, it strengthens the ecosystem, because the more data that stays inside, the harder it becomes to replicate that experience somewhere else.

This extends beyond software. If you use Google Home, Nest devices, Android Auto, Chromebooks, or Wear OS devices, the ecosystem expands into hardware as well. These products are designed to work best within Google’s environment. Outside of it, functionality often becomes limited. Even services like YouTube Premium or Google Workspace add another layer. Each additional service increases reliance on the system.

At this point, I could keep going. There’s a lot more to cover when it comes to Google alone. But that would take us out of scope for this piece.

If you’re interested in going deeper, I’ve covered Google in much more detail in my privacy series. Worth checking out if you want a clearer picture of what’s happening behind the scenes and how to manage your exposure better.

Industry-Wide Pattern

Google is just one example, and probably the most visible one, but this pattern exists across the industry.

Companies like Proton, Ente, Apple…etc. All of them either have a strong ecosystem or are building one in their own way. Some focus on privacy, others focus on hardware integration or start with a single product and expand outward.

Some companies handle this more responsibly than others and genuinely prioritize user interests, but the model itself remains the same because at the end of the day, ecosystems do two things very well:

Increasing convenience
Increasing dependency

The Cost of Staying

By this point, every interaction already feeds into the platform. That data is used to refine how the system behaves. On the surface, this feels like improvement. But it creates a dependency that isn’t easy to replicate elsewhere.

The same data that improves your experience is also used to extract value from it. Advertising becomes more precise. Features and pricing can be adjusted based on usage. Visibility is influenced by systems you don’t control.

This ties in directly with enshittification.

Anthropic’s recent trajectory demonstrates the pattern. The company signed a $200 million contract with the Pentagon in 2025, but made clear it did not want its technology used for mass surveillance of people in the United States or for fully autonomous weapons systems. According to BBC reporting, when Pentagon ordered Anthropic to allow unrestricted military use of its AI and Anthropic refused, the Pentagon retaliated by designating the company a “supply chain risk” and ending the contract. The ethical stance attracted users seeking an alternative to OpenAI that contributed towards a large user base boost for Anthropic.

The user base boost allowed anthropic to conduct different experiments. In April 2026, Anthropic updated its pricing page showing Claude Code as unavailable on the Pro plan. The company later stated this was only a test affecting 2% of new signups, though The Register noted the change appeared across all public-facing documentation and support materials. Anthropic reversed the change within hours after user backlash. The company maintains the test was limited in scope and not intended as a permanent policy shift.

The bigger picture is, once you’re locked in - the platform has the power to degrade service quality, change pricing structures, run A/B tests to determine willingness to pay, adjust data collection practices, or make the service shit*y because you are the one dependent on the platform.

The platform depends on you only as part of an aggregate user base. And individual complaints don’t carry weight when switching costs are high enough that most users won’t leave regardless of what changes.

What You Can Do

At this point, you should have a clear understanding of how these systems work. But the question is…what can you actually do about it?

First of all, it helps to check where you stand.

Signs of Dependency

Look at your current setup and answer honestly.

• Does losing access to one account affect your email, files, photos, and payments?

• Does the idea of exporting your data feel complicated or unfamiliar?

• Do most of your logins depend on “Sign in with” options from large platforms?

• Do you rely on default apps for search, maps, or storage without considering alternatives?

If the answer to even one of these is yes, then you’re likely more dependent on a system than you realize.

How to Reduce Lock-In

The goal here isn’t to completely avoid ecosystems. That’s neither realistic nor practical. The goal is to avoid depending entirely on one.

If you want a more unified setup without the same level of data collection, companies like Proton offer a reasonable alternative for core services like email, calendar, and storage. It is still an ecosystem, but the incentives are different compared to most large platforms. Even then, I still wouldn’t rely on it for everything.

A mix-and-match approach works better - use one service for communication, another for browsing, another for search. For example, you can keep Proton for email and storage, use browsers like Brave or Mullvad Browser, and rely on search engines like DuckDuckGo instead of defaults.

For platforms like YouTube, using a separate browser profile helps isolate activity instead of feeding everything into one account.

The same idea applies to hardware.

If you prefer a specific brand, that’s fine. Just avoid building your entire setup around it. The deeper the integration, the harder it becomes to step away later.

When choosing tools, a few things matter more than anything else.

Make sure your data can be exported easily. Prefer services that work across platforms instead of locking you into one environment. Where possible, choose tools built on open standards or open-source foundations.

Final Thoughts

An important thing to note - the article’s point wasn’t to paint every tech company as a problem. A lot of these products are genuinely useful. They solve real problems and make everyday tasks easier. That’s exactly why ecosystems work as well as they do. The convenience is real, and it’s understandable why people rely on it.

But it’s important to acknowledge that convenience comes with trade-offs.

The majority of people don’t really know how things behave once you’re inside the system. But if you understand how the system works, you’re able to make your own decision instead of defaulting into whatever is presented to you.

You can still use these tools, just remember to not rely on them for everything.

Thanks for reading.
See you next time.

One Last Thing

Nearly all people interact with systems without questioning them. This space is about doing the opposite aka understanding how things work, where they fail, and what that means in practice.

If that’s the kind of content you want more of, subscribe. And if this piece added something new to how you see things, pass it on. Good information only matters if it travels.

And if you’re active elsewhere, follow along on Bluesky or X (Twitter) or LinkedIn - those are the two places I show up most consistently outside Substack. A follow, a like, a repost - it matters to me, more than you think.

What You Might Like

• Ensh*ttification - How Platforms Turn Against You

• Privacy Fundamentals - Part 1: Authentication

• Privacy Fundamentals - Part 2: Communication

• Privacy Fundamentals - Part 3: Your Browser

There is a good chance you have seen a drone in the sky at some point. Maybe at a wedding, a concert, or an expensive photo shoot.

And there is an even better chance that the drone was made by DJI.

You might not know the name.

But if you have ever watched a travel video on YouTube, seen aerial footage in a movie or documentary, you have almost certainly encountered DJI’s products.

The company single-handedly controls somewhere between 70 and 90 percent of the global consumer drone market depending on which segment you look at.

A 2024 estimate puts DJI accounting for over 90% of all consumer drones sold globally.

More than the numbers, the interesting part is DJI’s story.

It starts with a boy losing his remote-controlled helicopter and figuring out why it crashed the way it did.

And the funny side to this story is that he initially didn’t plan to build this giant drone company.

So, in this article I’ll start with that story of how DJI came to be, how it stands against its competition and what the future looks like for consumer drones and DJI.

The Drone We’re Talking About

Before we go further, let’s make sure we are all on the same page about what a drone actually is and how it stays in the air.

A drone, in the context we are talking about, is basically a radio-controlled helicopter with four rotors. The pilot controls it remotely, either with a handheld controller or from a phone or tablet.

The kind of drones DJI makes are called multirotor drones. Most of the popular models are quadcopters, meaning they have four propellers arranged around a central body.

To understand how a drone stays up, think about what keeps a regular helicopter in the air. A single big rotor spins and generates lift.

The problem is that spinning a single rotor creates a rotational force that would spin the whole body of the helicopter in the opposite direction. That is why helicopters have a small tail rotor, to counteract that spin.

Drones solve this better. Two of the four propellers spin clockwise, and two spin counterclockwise. The opposing spins cancel each other out, so there is no spinning body.

To move forward, the drone tilts its front motors slightly down by spinning them a bit slower. To turn, it adjusts the speed of individual motors.

All of this happens automatically, dozens of times per second, through the flight controller chip inside the drone.

The flight controller is the brain of a drone. It reads data from sensors, processes it, and adjusts motor speeds constantly to keep the drone stable.

Without a flight controller, a drone would be nearly impossible for a human to fly manually.

This is the initial problem Frank Wang set out to solve.

Crash Of An RC Helicopter

Frank Wang is a guy born in China, in 1980.

As a child, he read a comic book in which the protagonist flew around the world in a small red helicopter. That image stuck with him and he wanted one of his own.

Our guy Frank studied hard enough that his parents eventually bought him a toy model helicopter. He crashed it the moment it left his hands.

That crash, instead of putting him off, made him obsessed. The problem was not the helicopter. The problem was control. Why was it so hard to keep these things stable in the air?

That question followed Frank Wang all the way to university.

He went to the Hong Kong University of Science and Technology. In his senior year, he took on a class project to build a flight control system for a model helicopter.

He stayed up nights working on it and presented the result to his professor, Li Zexiang.

The presentation was messy. The system did not fully work. But Professor Li saw something in Wang and brought him into the graduate program to keep developing it.

By January 2006, Wang and his small team had cracked it. The helicopter flew. It hovered. It landed. Stable and controlled.

Out of curiosity, Wang posted his result on a model airplane forum and offered the system for 50,000 Chinese yuan, which was about 7,000 US dollars at the time.

It costed him 2,000 dollars to produce it. To his surprise, people were willing to pay. And that’s when he saw a business.

In March 2006, he co-founded DJI with two teammates. The name stands for Da-Jiang Innovations, which loosely translates to Great Frontier Innovations.

His goal at the time was small. “If I can feed a team of 10 to 20 people with this, that would be great,” he said in an interview years later.

From A Dorm Room To The Skies

The early days of DJI were not glamorous.

Wang moved with his co-founders to Shenzhen after graduation, renting a small apartment and funding the operation with leftover scholarship money and support from family.

A family friend named Lu Di provided 90,000 US dollars and helped manage finances.

DJI’s first product was the XP flight controller series. These were not drones. They were the flight controllers that we just talked about above.

Essentially, flight controllers are brains that went inside drones that hobbyists were building themselves at the time.

Before companies like DJI existed, flying a radio-controlled aircraft was serious business.

You had to buy components from different sources, assemble them yourself, calibrate everything, and then pray it did not fall out of the sky.

Most people who tried lost aircraft in spectacular fashion.

The XP flight controller changed that for a lot of hobbyists. It came pre-built and provided stable, reliable performance out of the box.

DJI sold around 20 units a month, mostly through word of mouth and online model airplane forums.

In 2009, a team used DJI’s flight control system to fly a drone to the peak of Mount Everest. It was a proof of concept that the technology worked in extreme conditions. Orders started coming in more consistently after that.

Professor Li joined DJI as chairman. A family friend sold his apartment to invest in the company. A high school friend of Wang’s joined in 2010 to run marketing and began pushing DJI into hobbyist markets outside China.

The early years were hard. Wang was a perfectionist and an exacting boss. Most of the founding team eventually left.

There was even an incident where an early employee tried to steal DJI’s intellectual property and sell it to a competitor.

Wang wanted to sue but could not afford the legal fees. Professor Li helped settle the dispute and raised additional capital to keep the company alive.

Through all of this, Wang kept iterating on the technology.

The Drone That Made DJI

By 2012, DJI had built enough components around drones. They had the technology for flight controllers, motors, and GPS systems.

Wang decided to put it all together into a single product. And on January 7, 2013, DJI launched the Phantom 1, its first ever drone.

It was priced at 629 US dollars. It was white, compact, and required no assembly. You pulled it out of the box, charged the battery, connected a controller, and flew it.

It was the first consumer drone you could just fly. Full stop.

Every other quadcopter on the market was either a DIY kit or an industrial product costing tens of thousands of dollars.

It used GPS to hover in place automatically. If it lost signal from the controller, it would fly back to where it launched from on its own.

The press release DJI sent out was two paragraphs. It said: “You can fly your Phantom the moment you receive it.”

That year, DJI’s revenue went from about 4 million dollars to 130 million dollars.

The drone community had never seen anything like it. Filmmakers realized they could now get aerial shots that used to require hiring a helicopter crew at thousands of dollars per hour.

Wedding photographers (esp in our part of the world) realized they could shoot from angles no one had ever captured before. YouTubers could make their travel videos look cinematic too.

By 2015, DJI was the largest consumer drone company in the world. Competitors who had existed before the Phantom began quietly shutting down or pivoting away from hardware entirely.

But the Phantom was just the beginning of what DJI was building.

How DJI Builds Almost Everything Itself

One of the biggest reasons DJI got so far so fast is that it builds almost everything in-house.

Most technology companies buy components from other suppliers and assemble them into finished products.

DJI does some of that too, but the core technology and the software is designed and produced internally.

This is called vertical integration. It is the same concept as Apple producing their MacBooks and also developing the MacOS that goes inside it.

And this vertical integration is what has given DJI enormous advantages.

Wanna know how vertical integration helped Apple beat most of the laptop competition to dust? Here’s our coverage of the MacBook formula linked here.

Let’s break down what is actually inside a DJI drone and why each part matters.

The flight controller is the central computer of a drone. It reads data from the GPS module and many other sensors.

It processes all of that information and adjusts the motor speeds constantly to keep the drone level and on course.

Without a good flight controller, even a well-built drone handles like a car with no power steering.

The gimbal is the mechanical arm that holds the camera. The whole point of a drone is to get smooth, stable aerial footage.

But a drone is always vibrating, always tilting slightly in the wind. If you just strapped a camera to the body of a drone, your footage would look like it was shot during an earthquake.

A gimbal solves this by mounting the camera on a set of motorised arms that rotate in three axes: tilt, roll, and pan.

It reads the drone’s movement and adjusts in real time, keeping the camera perfectly level regardless of what the drone is doing.

DJI developed its Zenmuse gimbal series early on, and that became one of its key advantages.

Other companies trying to build camera drones had to source gimbals from third parties. DJI built its own, at a fraction of the cost.

The transmission system handles how video and control signals travel between the drone and the remote controller.

Here DJI developed its own proprietary transmission technology called OcuSync, which it later updated and renamed.

It allows stable video transmission at distances of several kilometres with low latency. If you are watching a live feed on your phone or controller screen while flying, that is the transmission system working.

The software is what ties it all together though.

The DJI Fly app runs on your phone and gives you a live view from the drone’s camera, flight telemetry, battery status, maps, and access to automated flight modes like ActiveTrack.

ActiveTrack lets the drone follow a moving subject automatically. DJI updates its app regularly and has released developer tools so third parties can build on top of its platform.

Alongside this, DJI makes everything from standalone camera stabilisers like the Ronin series, used by film crews on the ground, to the Osmo line of action cameras and smartphone gimbals.

These products share technology with the drone division and give DJI a broader presence in the creator tools market.

All of this is manufactured primarily in Shenzhen, which is often called China’s Silicon Valley.

Shenzhen has a dense ecosystem of component suppliers, electronics manufacturers, and engineering talent all found in one city.

If DJI needs a custom motor component, there is likely a supplier an hour away.

That proximity allows rapid prototyping and iteration at a speed Western companies simply cannot match with their supply chains.

DJI holds over 5,000 patents globally. About 25 percent of its more than 14,000 employees work in research and development.

From Hobbyists to Hollywood

In 2013 the Phantom made DJI famous among hobbyists. By 2016, the company was everywhere.

Hollywood adopted DJI drones rapidly. The cost of aerial cinematography dropped dramatically.

Film productions that used to budget tens of thousands of dollars for helicopter crews could now get comparable or better shots for a fraction of the price.

DJI won a Technology and Engineering Emmy Award in 2017 for its contributions to television production.

If you’re a fan of shows like Game of Thrones and Better Call Saul, just know that they used DJI’s technology at some point.

In 2016, DJI released the Mavic Pro. It was the first foldable consumer drone.

When folded, it was roughly the size of a water bottle. It had a four-axis gimbal, a 4K camera, obstacle avoidance sensors, and a flight time of about 27 minutes.

It could fly up to 7 kilometres from the controller.

The Mavic changed who could own a drone. You could throw it in a backpack and take it anywhere.

Travel creators went from filming everything at ground level to getting shots that looked like they were from a National Geographic documentary.

DJI then moved aggressively into the enterprise and professional market.

Its Matrice series drones are designed for industrial applications.

They have interchangeable payloads, which means you can swap out the camera for a thermal imaging sensor, a LiDAR mapping unit, a gas detector, or special cameras depending on what you need.

To learn more about LiDARs and how they’re useable in self-driving cars, here’s a linked article that we wrote a while back.

Firefighters use them to see heat sources through smoke. Energy companies use them to inspect power lines and wind turbines.

Construction firms use these drones to map building sites with centimetre-level accuracy.

The Agras line is DJI’s agricultural drone series. These drones are designed to spray pesticides and fertiliser over crops with precision.

In countries like China and Japan, where farmland is expensive and labour is costly, agricultural drones have saved significant amounts of time and money.

A single Agras drone can cover farmland far faster than a manual operator with a backpack sprayer. The system uses GPS mapping to ensure even coverage and avoids spraying areas outside the designated field.

In 2019, DJI acquired a majority stake in Hasselblad, the Swedish camera company known for its extremely high-quality image sensors.

Hasselblad cameras have been used on NASA missions and are considered among the best in the world for colour accuracy and image quality.

DJI began integrating Hasselblad sensors into its professional drones, including the Mavic 3 Pro and the Mavic 4 Pro released in May 2025, which ships with a 100-megapixel Hasselblad sensor and a 51-minute flight time.

The Ecosystem Nobody Can Resist

DJI’s dominance is not just about making good hardware. They do that too but the secret sauce is when that hardware combines into an ecosystem.

If you are a drone operator, your workflows are built around DJI’s software.

Your team knows how to use DJI’s controller and app. Your editing workflow uses DJI’s video formats. Your automated missions are set up in DJI’s enterprise platform.

Switching to a competitor is not just buying new hardware. It means retraining your team, rebuilding your workflows, and accepting that the new product probably does not do everything DJI’s product does, at least not yet.

DJI also offers developer tools and software development kits that let other companies build applications on top of DJI’s drones.

This created a whole industry of businesses whose products only work with DJI hardware. Drone inspection software, agricultural analytics platforms, mapping tools, all built assuming DJI is the platform.

This is similar to how Apple created an app store ecosystem that made iPhones stickier. You stay on the iPhone partly because your apps are there, your data is there, and your habits are formed around iOS.

recently covered Tech Ecosystems in a dedicated article which you can read here. You’ll understand why most companies are looking to build their own ecosystems.

DJI basically pulled off the same trick with drones.

Then there is the brand itself. For a large portion of the population, the word drone and the word DJI have become interchangeable.

In the same way people say Google when they mean search, people say DJI when they mean camera drones. That kind of brand recognition is built over years and is almost impossible to dislodge quickly.

Why The Competition Never Stood A Chance

Over the years, many companies have tried to compete with DJI. Most of them are no longer in the consumer drone business.

3D Robotics was one of the most prominent early competitors. It was a US-based drone company that raised significant investments.

By 2016 it had essentially exited the consumer drone hardware market because it could not compete with DJI on price or features.

The founder himself said: “I don’t see a way to compete with DJI on consumer hardware.”

GoPro tried too. The company known for its action cameras launched the GoPro Karma drone in October 2016, the same month DJI launched the Mavic Pro.

The Karma was recalled weeks after launch due to a battery issue that caused some units to lose power mid-flight. GoPro exited the drone market entirely by early 2018.

Skydio is the most interesting current competitor.

It is a US company that focuses on AI and autonomous obstacle avoidance. Its drones are genuinely excellent at following a subject through complex environments without crashing.

However Skydio drones cost significantly more than equivalent DJI models and their camera quality has historically not matched DJI’s.

Skydio raised 170 million dollars in late 2024 and has secured government and military contracts, but its consumer market share remains very small.

The fundamental problem for competitors comes down to a few things DJI does that are very difficult to replicate at the same time.

DJI releases new products very fast because it controls its own supply chain and does much of its manufacturing internally.

It can go from prototype to shipping product faster than most Western companies can go from prototype to first supplier meeting.

DJI prices its products aggressively because it manufactures at massive scale in China with vertically integrated components.

Its cost per unit is lower than what competitors can achieve by buying components from different suppliers.

DJI invests a lot in the user experience too.

The DJI Fly app, the return-to-home feature, obstacle avoidance, ActiveTrack, and all the intelligent flight modes are things that matter enormously to a first-time drone buyer.

A drone that is technically impressive but hard to use will not sell. DJI understood this from the moment the Phantom launched.

DJI also had the first-mover advantage. It built an ecosystem before anyone else understood what the market would become.

Geopolitics Enter The Chat

DJI’s dominance has created a serious geopolitical problem, at least from the POV of the US government.

Their concerns are straightforward. DJI drones collect data. They record video of wherever they fly, GPS coordinates, altitude, flight paths, and more.

The question governments have asked is: where does that data go, and who can access it?

The US government began raising concerns formally in 2017, warning about surveillance risks from Chinese-made drones.

In 2019, the US Army banned the use of DJI drones entirely.

In 2020, the US Commerce Department placed DJI on its government’s economic blacklist for its alleged role in a surveillance project which DJI denies.

The ban has not destroyed DJI globally.

The US market is important but it is not the whole world. DJI has been growing in other markets and its global market position remains largely intact.

But it does create a real ceiling on its ambitions in the world’s largest tech economy. DJI continues to operate normally outside the United States.

For us in Pakistan and most of the rest of the world, none of this directly affects the ability to buy or fly DJI products. We couldn’t buy them even before without breaking a bone :p

Where DJI Goes From Here?

DJI is not slowing down anytime soon.

In May 2025, it released the Mavic 4 Pro with a 100-megapixel Hasselblad sensor, a 360-degree rotating gimbal, 51 minutes of flight time, and obstacle avoidance in every direction.

It is a genuinely impressive piece of technology. But DJI is also moving into areas beyond consumer drones.

Artificial intelligence is a growing part of what DJI’s drones can do.

The obstacle avoidance systems on newer models are sophisticated enough that even a new pilot can fly through a complex environment without crashing.

ActiveTrack has gotten good enough that filmmakers use it without a dedicated drone operator.

The drone watches a moving subject and follows it autonomously while the operator focuses on framing and story.

DJI is also investing in autonomous infrastructure.

The DJI Dock 2 is a charging station that lets enterprise drones take off, complete a mission, land, recharge, and take off again without any human intervention.

This means a drone can monitor a location continuously without anyone physically present.

The FlyCart 30 is DJI’s drone delivery product, designed to carry payloads of up to 30 kilograms.

It is aimed at deliveries in remote or difficult-to-reach areas, supplying medicine, supplies, or equipment where traditional transport is slow or expensive.

DJI has also begun moving into self-driving technology.

It has a LiDAR and autonomy division that sells sensors and software to electric vehicle manufacturers for use in assisted driving systems.

This is diversification that takes DJI well beyond drones entirely.

In 2024, DJI began producing electric bicycle motors. In 2025, it released the DJI Romo, its first robot vacuum cleaner.

These expansions show that DJI sees itself as a broader robotics and autonomy platform, not just a drone maker.

The company Frank Wang he built in a Shenzhen apartment is now one of the most technically sophisticated consumer hardware companies in the world.

These guys started with simple flight controllers to drones and full-blown ecosystems and now they’re diversifying that too by going into autonomous driving, electric bicycles even when Uncle Sam has pulled his blessings.

For me personally, it’d be really cool to try DJI’s products here in my neighborhood but from what I hear it’s a pain to get hold of consumer drones in Pakistan.

They’re not only expensive and imported through grey import but even if you are able to get one, flying it is definitely a challenge that involves law enforcement agencies.

Still, even if I couldn’t buy or fly these drones today, I can certainly write for you guys and share how reading about the technology excites me. Maybe someday I might be able to fly one and write a hands-on review here on SK NEXUS.

Comment down below if you’re excited for that ;)

Similar Articles from SK NEXUS

Further Learning

• The Consumer Electronics Hall of Fame: DJI Phantom Drone

• Bad Karma: GoPro’s first drone does not impress

• GoPro quits the drone business

• Why China’s dominance in commercial drones has become a global security matter

• DJI still dominates the 2025 drone market

• DJI Developers Share the Secret

• DJI vertical integration, 5,000 patents

• DJI acquires majority stake in Hasselblad

• DJI is skipping the US with its most advanced drone yet

• Review: DJI Mavic 4 Pro

• DJI ban: how the world’s biggest drone maker is getting shoved out of the US

• And the Emmy goes to… Apple, DJI, Microsoft

So, I was scrolling through YouTube the other day and came across an old video by Marques Brownlee on Apple’s ecosystem and it got me thinking…how do we actually end up using the products we use?

I bet none of you just sat down and decided - “I’m going all-in on this ecosystem.” It doesn’t work like that. You buy a phone, then maybe a laptop. Then something else that just works with it and before you know it, you’re inside a system.

But does all this happen by itself? Probably not.

Tech companies focus on building connected environments where everything is designed to work together and feel smooth. But behind that smooth experience, there’s a structure shaping how you interact with everything.

Let’s take a closer look at that.

What an Ecosystem Actually Means

At its core, an ecosystem is when a company designs its hardware and software to function as a single, connected experience. The value doesn’t just come from one device or one app. It comes from how everything works together.

Whether it’s your phone syncing with your laptop or the messages showing up across devices or even the files moving without you thinking about it - that’s the ecosystem doing its job.

Take the Apple Inc. ecosystem as an example. Phones, laptops, watches, services, subscriptions…etc. everything is built to connect without friction. It’s one of the most polished implementations of the ecosystem out there, and that’s exactly why it works so well.

How Ecosystems Are Built

Frankly, the ecosystem - like everything in life - doesn’t just pop out of nowhere. There is a whole strategy consisting of multiple steps behind it:

Step 1 - The Entry Point

It all begins with a product that solves a clear problem.

This could be a phone, an email service, or even a search tool.

The companies craft such a product which is easy to adapt aka The product works pretty well on its own & is easy to start using. And once it becomes part of daily use, it forms the base for everything that follows.

Step 2 - Expansion

Once the initial product gains traction, more services start appearing around it - Storage, messaging, backups, and other tools begin to extend the original experience. Each addition makes the system more useful and keeps more of your activity inside it.

Every moment you spend inside the app, starts to feel better because features that were once missing are now there and the overall experience is just better.

Step 3 - Integration

At this stage, products begin to work together.

For example - your devices start to sync automatically and certain features only work when multiple products are used together…etc. In other words, the experience becomes consistent across devices and it feels like you are operating inside a connected setup.

This also starts influencing decisions. Suppose if you already use one product, choosing another from the same system feels easier.

Step 4 - Dependency

Over time, the system becomes part of routine use because everything is perfectly connected together.

Leaving remains possible, but it requires effort because data needs to be moved and habits need to be adjusted. All of this requires a person to step out of their comfort zone and majority people don’t like to do that pretty often.

This is often referred to as lock-in.

Now, don’t get me wrong - tech companies don’t literally put some steel collar on you, but a digital steel collar which is enforced through convenience? Yes.

This connects with the enshittification pattern discussed in my other article as well. I talked about how once the users are settled inside a system, the focus moves from adoption to retention. Worth checking out that piece as well.

Case Studies

The companies listed here aren’t the only ones building ecosystems. Almost every major tech company follows a similar approach in some form.

The reason for focusing on a few names is because larger systems make the pattern easier to observe.

Apple Inc

Apple Inc. is one of the clearest examples of a tightly integrated ecosystem.

Its approach is based on controlling both hardware and software. This allows it to maintain consistency across devices. Products like the iPhone, Mac, Apple Watch, and AirPods are designed to work as parts of a connected system rather than standalone tools.

Features such as iMessage, AirDrop, FaceTime, and iCloud sit at the center of this setup. They connect devices and keep data moving across them.

This level of integration is what makes the experience feel seamless. As also discussed in content by Marques Brownlee, the system works well because everything is designed for that environment.

At the same time, this design introduces limits.

Devices often assume that other Apple products and services are already in use. For example, products like the Apple Watch depend heavily on the iPhone to function fully. Outside that setup, their usefulness drops.

Another aspect is scale.

Apple’s ecosystem is not just large in design, but also in reach. A single product line like AirPods generates tens of billions in revenue annually, with estimates placing it in the range of roughly $18–20 billion in recent years. That is comparable to the size of entire standalone companies.

This gives a sense of how deeply these products are embedded into everyday use.

Over time, moving away from the system becomes more complex because as discussed earlier - users are not only switching devices. They are moving away from stored data, app purchases, communication history, and features that connect across devices.

The friction comes from how many parts of daily use are tied together and this does not make the ecosystem inherently negative. It highlights the trade-off. A more connected experience brings more dependence on the same system.

Google

Google, on the other hand, takes a different route, but the outcome is similar.

Instead of starting with hardware, its ecosystem grew from software. Search, Gmail, Google Drive, YouTube, and Android form the base. All of these are tied together through a single Google account, which allows data and activity to move across devices.

Over time, hardware was introduced as well. Pixel phones, Chromebooks and wearables extended the system further.

Compared to Apple, this setup is more flexible. It works across platforms like iOS and Windows. But flexibility doesn’t mean independence. Once you’re logged into a Google account, a large part of your digital activity sits in one place and this kind of integration becomes stronger over time because services are designed to depend on each other.

A good example of this is the Play Store.

On Android, it’s the default source for apps. Alternatives exist, but they don’t offer the same range or reliability. Nearly all users stick with what’s already there.

There’s also another layer.

Many apps rely on Google Play services to function properly. Notifications, location, payments, and other features often depend on it. Even if you install apps from outside the Play Store, parts of them still connect back to Google’s infrastructure.

This makes moving away from the ecosystem harder than it seems.

So while Google’s system feels more open on the surface, it still builds the same kind of dependency over time, just through a different approach.

Why Ecosystems Work So Well

At this point, it’s clear that ecosystems are everywhere. Large companies have them and smaller ones are trying to build them.

The question is why they work so well.

Here’s a hint - convenience is part of it, but it’s not the only reason.

Let’s take a look at the other reasons.

1 - Connected Features

Ecosystems create features that only work when multiple parts are used together. - Services like cross-device messaging, instant file sharing, synced notifications, or shared clipboards depend on this connection.

Individually, these tools might not seem like a big deal but together, they offer something harder to replicate using separate services.

There’s also a behavioral side to it - if using one more device unlocks additional features, an average Joe won’t think twice. He’ll just add it and that’s exactly how the system expands.

2 - Less Choice, Less Effort

Take streaming platforms - you open Netflix or Hulu and get hit with endless options due to which you spend more time deciding than actually watching.

More choices don’t always help and ecosystems reduce that which is why instead of choosing between multiple software or hardware devices - people just end up going for a company that provides an ecosystem which has all the necessary hardware devices or services because that’s what an average Joe needs.

3 - Staying Becomes Easy

When everything is all set up - the go-to decision becomes staying where you are and that means continuing using what already works because switching requires effort and staying doesn’t.

And over time, that turns the ecosystem into the environment where everything happens aka work, communication, storage, daily tasks…etc.

Why? Because it became the easiest option.

The Tradeoffs

The same structure that makes ecosystems work well also creates trade-offs.

Flexibility decreases with the increase of tools. That means switching to another service or device means losing certain features or rebuilding how things work which takes a lot of effort.

And this isn’t about companies trying to make things worse for users, rather it’s more about how systems are designed in a way that makes leaving harder than staying.

There have also been cases where even basic actions were made difficult.

For example, the Federal Trade Commission filed a case against Amazon over its subscription practices, where canceling Prime was described as unnecessarily complicated.

Situations like this show how friction can be built into systems and once you look at it from that angle, you start to see how a more connected experience increases dependence on the same system.

Final Thoughts - Don’t Put Everything in One System

Ecosystems are not inherently good or bad.

They exist because they solve real problems. They make technology easier to use. For the majority of people, they provide real value and voiding them completely isn’t realistic.

The real question is how much you rely on a single system.

Keeping everything in one place makes things simple in the short term. Over time, it also increases dependence. Spreading tools across different systems adds some friction, but it also gives flexibility.

I believe that technology is moving toward more connected systems and understanding how they work is the only thing that will allow you to use them, instead of being used by them.

Make sure to stay tuned for the second part of the tech ecosystem series in which we will be diving much deeper into the system.

One Last Thing

Nearly all people interact with systems without questioning them. This space is about doing the opposite aka understanding how things work, where they fail, and what that means in practice.

If that’s the kind of content you want more of, subscribe. And if this piece added something new to how you see things, pass it on. Good information only matters if it travels.

What You Might Like

• The Truth About the Dark Web: Myths vs Material

• How We Got Online: The Hidden History of the Internet

• What are LLMs & How Are They Changing Our World?

• Blockchain - The Cool Parts of the Tech Behind

If you’re into PC gaming these days you might have gone over headlines calling for Intel’s comeback.

And honestly, I get why you might be skeptical. Intel has been saying it’s coming back for years now and most of the time it was just corpospeak.

But this time something actually feels different.

See, Intel just launched their latest desktop processors and once you go over their specs and the price, you have to give them a hand. It’s exciting news from Intel since quite some time.

So let me take you through the whole story. Not just the chips that launched last week, but how we got here.

In today’s article I’ll share how Intel went from owning the chip market to becoming a meme, how AMD quietly took over, and now what these new chips actually mean for anyone buying a PC in 2026.

I’ll also talk about something that could be even bigger. NVIDIA making CPUs (Yes, the GPU company). It could be the most interesting thing to happen to PC hardware in a long time.

But first, let me explain what a CPU actually is because not everyone knows and that is completely fine.

What Is A CPU Anyways?

Before I even start let’s just establish that the words ‘processor’ and ‘CPU’ stand for the same thing.

In any computer, CPU is the brains of the operation, a tiny chip that does billions of small mathematical calulations in a second to get you to browse the internet or play games or whatever you do on your computer

A CPU generally dictates how fast and powerful your computer would be. Your PCs, smartphones, laptops - basically every computer needs a processor to function properly.

If you want to understand CPUs at a more deeper level, you can checkout my full blogs where I dive deep into Cores, Threads and what you should know when selecting a CPU for your PC. That article is linked here.

The x86 Duopoly

For about two decades now only two companies have been making CPUs for regular consumers using PCs. Those are Intel and AMD.

And it’s not because other companies cannot make CPUs for PCs but rather both Intel and AMD keep a secret to them that they don’t share with others.

That secret is the foundation of the processors both companies make. That is the x86 instruction-set architecture.

Think of an instruction-set architecture (ISA) like an underlying blueprint or manual for making processors.

At this point, x86 is decades old but only Intel and AMD have a license to it. It’s age is the reason it is the dominant platform in desktop PCs even to this day.

Having access to x86 makes Intel and AMD sort of a monopoly (duopoly to be exact) because only these two companies have license to it.

Even if x86 is decades-old, it is being challenged by a new player: ARM. It is the one your MacBooks use.

If you’re curious about ARM, I’ve talked about it too here on SK NEXUS, you can read the full article here.

That is partly why historically there has been little to no competition in the CPU market. The usual trend is that the dominant player gets lazy with time and stops innovating.

However this is changing because a new player is entering the ring. I’ve talked more about it later in the NVIDIA section.

Intel Before Ryzen

For a long time, Intel has been the dominant CPU maker in the desktop and enterprise computer markets. The company used to own the PC world.

It was called ‘Chipzilla’ for its market dominance because it single-handedly controlled somewhere around 80 to 90 percent of the CPU market for decades.

However, Intel’s insane success made it lazy.

If you bought an Intel CPU in 2015 and then looked at what Intel was selling in 2017, the improvement was honestly pretty borderline.

Just to give you an idea how lazy Intel got, I wrote a deep-dive article discussing the major issues with Intel that made it lose the ‘Chipzilla’ title, the deep-dive article on Intel is linked here.

Intel was stagnating from the technology side in the mid 2015-era and around that time AMD showed up with a punch to Intel’s face.

AMD’s Ryzen Changed Everything

In early 2017, AMD released its Ryzen processors. Those were the first generation processors on their new architecture they called Zen.

Ryzen brought something Intel wasn’t offering at those prices: more cores, more threads, better multitasking, and a lower price tag.

The second generation Ryzen processors (Zen 2) in 2019 were a big leap forward. Third generation Ryzen processors (Zen 3) in 2020 finally took the gaming crown away from Intel.

By 2020, recommending Intel over AMD for most PC builds was genuinely hard to justify.

AMD had more than doubled its market share in the desktop segment.

Intel tried to respond by stuffing more cores into its chips and pushing clock speeds higher. But efficiency was still not there.

The chips ran hot, used a lot of power, and fell behind in too many comparisons.

And that’s when things started to get genuinely bad for Intel.

Intel’s Dark Years

From around 2019 through 2024, Intel was struggling real hard.

But the worst part of this whole era was what happened with Intel’s 13th and 14th generation processors.

People randomly started noticing that their latest-gen Intel CPUs were quietly dying. Not just slowing down or underperforming. Physically, permanently degrading over time.

The 13th and 14th generation i9 processors (Intel’s top consumer chips) were the main culprits. Systems would crash. Games would fail. Blue screens of Death became a daily thing for some users.

What was happening is that these chips were being pushed with too much voltage for too long. This was causing permanent physical damage to the intel inside (pun intended).

And the worst thing was that this damage could not be reversed by any software update. Intel first tried to shift blame to motherboard manufacturers. Then it released patches. Multiple rounds of them across 2024.

Eventually Intel admitted there was a fundamental firmware-level issue with how the chips handled voltage, extended the warranty on affected chips to five years, and told owners to update their BIOS.

The update patches could prevent future degradation on newer chips but they could not undo damage that was already done. If your chip had already degraded, it stayed degraded.

And that is when consumers started losing trust.

AMD had none of these issues during this period. Ryzen chips just worked. Stable, efficient, reliable.

AMD’s market share was also growing.

Intel went from being the leader in chip manufacturing to a company looking to just survive especially after Ryzen gave it a punch to the face.

We covered Intel’s decline in a lot more detail starting from the time they missed the smartphone wave and lost Apple as their customer.

If you’re interested in learning more about Intel’s issues, here’s our deep-dive article which you can read here.

Intel’s 2026 Comeback

This March, Intel launched two new desktop chips:

1. Core Ultra 7 270K Plus at $299

2. Core Ultra 5 250K Plus at $199

Keeping the weird naming scheme aside, the chips are actually impressive. LinusTechTips also titled their review video as ‘Intel is back’.

GamersNexus reviewed the 270K Plus and called it “Not a Waste of Sand.” And that’s coming from a reviewer who termed the previous generation Intel processors as ‘A Waste of Sand’ :p.

Tom’s Guide also said Team Blue is getting its desktop swagger back.

So you might be asking: What actually makes these chips different from what came before?

The 270K Plus is a 24 cores beast with 8 performance cores and 16 efficiency cores.

For $300. That is the same core configuration as Intel’s previous flagship chip, the 285K, which launched at over $600.

Intel basically figured out how to offer its best silicon at half the original price.

In multi-threaded workloads, so things like video rendering, photo editing in batch, 3D work, compiling code, the 270K Plus beats AMD’s Ryzen 7 9700X by over 100 percent.

TechSpot showed the 250K Plus was 85 percent faster than AMD’s Ryzen 5 9600X in Cinebench multi-thread. Both Intel chips cost the same as their AMD competition.

To be clear, multi-threaded work is the stuff that uses many cores at once. It’s what content creators, editors, and developers care most about, not our average Bashir. But Intel is now dominating at this price.

Intel also launched a driver suite with these new processors. Think of it like a software driver that handles advanced optimizations for your processor.

It’s Intel’s equivalent of what AMD has had for its Ryzen platform for a while. A bit late to the party but it does the job.

Is Intel Better Than AMD Now?

Honestly, this is where the honest answer is not as straightforward as you’d want it to be.

In productivity workloads (general tasks like browsing, content consumption) at the $300 price point, yes, Intel is genuinely better right now. AMD has nothing at $300 that matches the 270K Plus even in rendering and similar tasks. Nothing close.

In gaming however, AMD still has the edge. Especially with their X3D chips.

The Ryzen 7 9800X3D costs around $480 and it is the best gaming CPU you can buy right now. It beats the 270K Plus in most gaming benchmarks. But it costs $180 more.

The 270K Plus is not that far behind in gaming though. It’s competitive. And for someone who uses their PC for a mix of gaming and productivity work, the Intel option at $300 suddenly makes a lot of sense.

A review described the 270K Plus as a “corker of a comeback for Intel, particularly in the realm of productivity.”

The value picture is genuinely different from what it was a year ago.

Intel is now the chip you would recommend to productivity-heavy users at $300. AMD is still the chip you recommend for pure gaming if budget allows.

Just look how the tables have turned and we’re seeing Intel be the value brand.

Not A Bed Of Roses

There is something important I need to tell you before anyone runs out to buy one of these chips.

The motherboard these Intel chips launched on, called LGA 1851, is a dead-end platform. Intel has already said it’s moving to a new socket with its next generation of processors.

What this means practically: if you buy a 270K Plus today and want to upgrade your CPU two or three years from now without buying a new motherboard, you cannot. The next Intel generation will need a new motherboard.

AMD’s current motherboard platform, AM5, still has more life left in it. AMD has promised at least one more CPU generation on AM5. So AMD users can upgrade the CPU later without replacing the whole motherboard.

This matters if you are someone who upgrades bit by bit. If you buy a chip and use it for five or six years without touching it, the platform situation barely affects you.

The 270K Plus is strong enough to last that long. But if you like swapping parts every couple of years, this is worth knowing upfront.

GamersNexus said it clearly in their review: “If Intel can repeat a launch like the 270K Plus but on a platform with some actual life to it, they’d be negating the number one complaint.”

Also worth noting: early reviews showed the chip running slightly below its advertised clock speeds in some configurations. Not dramatically but reviewers did flag the small differences.

And there is always the question of whether the listed prices hold once the chips actually land on shelves at scale.

And honestly that is the biggest concern. Because of AI we’re already in a crisis- like situation and then there are two major wars happening just before us.

Supply issues can push prices above MSRP quickly. And by the time this stuff gets to our local shopkeepers we’re likely to see a significant price increase which is just sad.

NVIDIA Joining The CPU Chat

Now here is something no-one saw coming. There have been rumors that NVIDIA is going to launch its laptop CPUs.

Yup, that graphics cards maker which just hit 5 trillion dollar market cap. The same people behind your RTX GPU.

Wondering how a GPU company just became the most valuable company in the world? Then, you may checkout my recent article on NVIDIA that answers this burning question that many people have.

They are making their own processor chips called the N1 and N1X and these are confirmed for Dell and Lenovo laptops that we may see somewhere around next year.

A Dell 16 Premium laptop with an N1X chip inside was spotted in a supply chain document from late 2025. Multiple OEMs are reportedly preparing products around this processor.

What makes this interesting is what the chip is reportedly capable of.

The N1X is expected to pack GPU performance in the ballpark of an RTX 5070/5060 TI, take a second and read that again, RTX 5070 performance by an integrated chip!!

If you read our MacBook Neo article, you already know what happens when a company puts CPU, GPU, and RAM all within one chip using a shared architecture.

Performance goes up. Power consumption goes down. Battery life improves dramatically. The machines become thin and quiet. That is what NVIDIA is attempting for Windows laptops.

The gaming implications alone are significant. Right now a gaming laptop needs a dedicated GPU which adds cost, weight, and heat. Just try and use a gaming laptop and an ultrabook side-by-side.

An N1X laptop with integrated RTX 5070 class performance could in theory give you serious gaming in a machine that looks and feels like a normal thin laptop.

Though I’ll be honest there are still real questions.

Windows on ARM is not perfect yet. App compatibility has improved a lot since the early days but gaming on ARM is largely unheard of.

The N1X has reportedly hit some engineering delays and may not land until late 2026 or early 2027. However there are reports that Dell and Lenovo are up for a reveal at Computex 2026 coming up in June.

However, it may be some time until we see products with these chips become available for purchase.

But the direction of travel is clear. NVIDIA entering this space forces Intel and AMD to think differently about what an integrated chip can do.

Competition Gave Us Better Processors

If you’re a regular reader of SK NEXUS you’d have noticed a trend of us shouting that competition in products is a good thing for the end-user. And that’s something we really believe in.

At the end of the day, it’s a bloody competition that delivers great products.

If Ryzen didn’t give a solid punch to Intel we’d not have seen processors like the 270K and 250K Plus from Intel certainly not at this price.

When Ryzen came to the scene Intel had to improve almost immediately. It added more cores to its chips for the first time in years. Prices dropped. Performance jumped faster than it had in a long time.

That is what competition does to a market.

The 270K Plus at $299 offering double the multi-threaded performance of a same-priced AMD chip is a direct result of Intel needing to fight again.

A year ago that chip would have cost $600. Competition brought the price down.

So, no matter how big of a ‘Chipzilla’ you are, the product is what matters in the end.

The CPU wars being back is good news. Not just for people who enjoy reading nerding out on specs but those who are on a budget and want a good value for their money.

More competition, better products, lower prices. That is the story here. And honestly, it is only going to get more interesting from here now that NVIDIA is joining the chat.

As always your comments are what I look forward to when writing these articles. Your feedback directly helps us become better writers so please be generous with that in the comments down below.

Further Learning

• Not a Waste of Sand: Ultra 7 270K Plus CPU Review & Benchmarks

• Intel is BACK. THIS IS NOT A DRILL

• NVIDIA N1X and N1 CPU: Everything we know so far

• NVIDIA N1/N1X SoC surface on laptop motherboard

• Intel finally confirms instability issues with 13th-gen

• A timeline of the rise and decline of Intel

• With an Intel recovery underway, all eyes turn to its foundry business

Artificial intelligence is now everywhere. The term appears in product launches, marketing campaigns and everyday conversations about technology. Yet much of what happens behind the scenes remains poorly understood.

In 2025, a small controversy on YouTube highlighted this confusion.

The platform was quietly testing a feature designed to enhance video quality on Shorts. The system used algorithms to reduce blur and improve clarity during processing techniques similar to those already used in modern smartphone cameras.

But, the creators didn’t like it and quickly pushed back.

Many worried the platform was introducing AI slop into their content without consent, assuming the feature relied on generative AI to alter or recreate their videos.

YouTube responded publicly:

“No GenAI, no upscaling. We’re running an experiment on select YouTube Shorts that uses traditional machine learning technology to unblur, denoise, and improve clarity in videos during processing.”

The company emphasized that the experiment relied on conventional machine learning methods rather than generative AI.

For many observers, the explanation raised an obvious question.

“Aren’t those the same thing?”

They are not.

And that confusion points to a larger problem in how artificial intelligence is discussed today. Terms like machine learning, deep learning, generative AI, and even broader ideas such as AGI and ANI rarely make it into public conversation. Instead, the label “AI” is slapped on top of them. The result is a blurred understanding of what these systems actually are and how they work.

This article focuses on clearing up that confusion.

The Foundation: What Machine Learning Actually is

Machine learning (ML) is a branch of artificial intelligence focused on building systems that learn from data rather than following fixed instructions. Instead of programming every rule manually, developers train algorithms on datasets so the system can recognize patterns and make decisions on its own.

In other words, machine learning allows computers to classify information, detect patterns, and make predictions based on what they have already seen.

Take phishing emails for example - it might occasionally slip into your inbox, but the system quickly learns from patterns shared across similar messages and as more data is processed, the filter improves its ability to recognize and block those messages automatically.

The concept itself is not new. Researchers have been exploring machine learning since the 1950s, long before the current wave of AI hype. Today, many everyday systems rely on it, including spam filters, recommendation engines, tools that upscale video quality and much more.

Most of these systems do not generate new content. Their role is to analyze existing data and make informed decisions based on patterns within it.

Self-driving vehicles are another great example. These systems process enormous amounts of sensor and camera data to recognize road signs and navigate traffic. The model improves by learning from previous driving data and continuously refining how it responds to real-world situations.

In many ways, machine learning forms the technical foundation beneath much of what we now describe as AI.

Machine learning itself comes in several forms. While the field is broad, most applications fall into three major categories.

Supervised Machine Learning

Supervised learning involves training a model using labeled data, where both the inputs and the correct outputs are already known. The algorithm learns how to map inputs to the correct results by studying these examples.

For instance, a model can be trained using labeled images of animals. Each image already includes the correct label. After training on enough examples, the system learns the visual features associated with each category. When it encounters a new image, it can predict which animal it contains.

Unsupervised Machine Learning

Unsupervised learning works differently. Instead of labeled data, the model receives raw information without predefined answers and must identify patterns on its own.

A common example appears in customer data analysis. If a company has large datasets about purchasing behavior but no predefined categories, an unsupervised model can group customers with similar habits. These clusters can then help businesses understand more about each customer and then develop marketing strategies accordingly.

Reinforcement Learning

Reinforcement learning trains systems through trial and error. Instead of labeled examples, the model interacts with an environment and receives feedback in the form of rewards or penalties.

Consider an AI system learning to play chess. When it makes strong moves, it receives positive feedback and vice versa. Over time, by experimenting with different strategies and observing the results, the system learns how to make increasingly effective decisions.

Where the Confusion Begins: Traditional ML vs Generative AI

This distinction is where the YouTube controversy started to get messy. People assumed the feature involved generative AI, when the technology behind it was much closer to the machine learning systems that have existed for decades.

Traditional Machine Learning

Earlier we discussed machine learning as the broader concept and within it, traditional machine learning handles the “analyzing data and making predictions part” which I previously explained in detail.

Just keep this in mind - traditional ML works with existing data and tries to determine what something is or what might happen next and it does NOT generate new material.

Generative AI

Generative AI follows a different approach.

Instead of focusing only on prediction or classification, generative models are trained to produce new outputs that resemble the data they were trained on. The model studies patterns across massive datasets and learns how to recreate similar structures.

This ability allows generative systems to produce things like written text, digital images, music, computer code, or even video.

Tools such as ChatGPT generate natural-language responses. Image generators like Midjourney produce artwork from simple prompts, while emerging video models such as Sora can create entire scenes from text descriptions.

Deep Learning and the Architecture Behind Modern AI

Beyond machine learning and generative AI, two technical ideas power much of the AI people interact with today - deep learning and transformer models.

These systems are rarely discussed outside technical circles, yet they sit at the core of many modern AI advancements. Together, these technologies power many of the systems that define modern AI.

Understanding Deep Learning

Deep learning is a branch of machine learning that uses artificial neural networks with many layers to analyze data.

The relationship between these fields is often easiest to picture as a hierarchy. Artificial intelligence sits at the top, machine learning forms a major subfield within it, and deep learning represents a more specialized area inside machine learning.

The key difference between traditional machine learning and deep learning lies in how systems learn patterns.

Traditional ML systems often rely on human engineers to define the features a model should examine. For example, if a model needs to identify stop signs, programmers may specify characteristics such as shape, color, or edge patterns.

Deep learning models reduce this manual work. By analyzing large datasets, neural networks can learn those characteristics automatically. Instead of being told what features matter, the system discovers them during training.

This ability to learn features automatically allows deep learning models to handle far more complex tasks. Today they power technologies ranging from computer vision systems and speech recognition to robotics and self-driving vehicles.

Understanding Transformer Models

In 2017, researchers introduced a new neural network architecture called the transformer. This design quickly became the foundation for many of today’s most powerful AI systems.

Transformers are built to process sequences of information, such as sentences, lines of code, audio signals, or even frames in a video. Their key innovation is a mechanism known as attention, which allows the model to evaluate how different parts of an input relate to each other.

Instead of reading data strictly in order, transformers can analyze relationships between all elements at once. This helps the model understand context far more effectively.

Although transformers are often associated with text generation, the architecture is not limited to language. Variants such as Vision Transformers (ViT) are used in computer vision, while similar approaches power multimodal models that combine text, images, and audio.

Many modern AI systems rely on this architecture. Models such as ChatGPT and Claude are built on transformer-based designs, typically using encoder–decoder or decoder-only variants.

The Three Levels of AI

When people talk about the future of AI, they usually refer to three stages: narrow intelligence, general intelligence, and superintelligence.

Artificial Narrow Intelligence (ANI)

Artificial Narrow Intelligence is the only form of AI currently in real-world use. The word narrow simply means specialization. These systems are built to perform specific tasks extremely well. They operate within defined boundaries and do not extend beyond them.

Everyday technologies fall into this category. Voice assistants respond to commands, recommendation systems analyze user behavior, and tools like ChatGPT generate text or code. Each system performs its function with somewhat impressive accuracy, but none of them truly understand the world beyond their training.

And of course, the limitations exist, for example - A system trained to recommend movies cannot suddenly start driving a car or managing finances.

Artificial General Intelligence (AGI)

Artificial General Intelligence represents a system that can think, learn, and adapt across multiple domains, much like a human.

Unlike ANI, an AGI system would not be limited to a single task. Theoretically, it could possibly learn new skills, apply knowledge across different fields, and solve unfamiliar problems without retraining.

Despite growing public discussion, AGI has not been achieved. Headlines suggesting systems becoming conscious or achieving human-level intelligence remain speculative and often driven by hype rather than technical reality.

Artificial Superintelligence (ASI)

Artificial Superintelligence goes a step further. It describes a system that surpasses human intelligence, reasoning, creativity, decision-making, and even social understanding.

In theory, an ASI system would be able to operate at a scale far beyond human thinking.

The idea often appears in science fiction because of its extreme implications. A system capable of compressing decades of scientific progress into seconds would fundamentally reshape every field it touches.

It’s important to note that - at present, ASI remains entirely theoretical. Even reaching AGI is still an unsolved challenge, which places superintelligence far beyond current capabilities.

Why Everything is Suddenly AI

“If most systems are just machine learning, why does everything get labeled AI?”

It’s because of moneyyy!

A useful comparison comes from the term greenwashing, where companies exaggerate how environmentally friendly a product is to appeal to consumers. A similar pattern is now visible in tech, often referred to as AI washing - the practice of overstating products as “AI-powered” to make them appear more advanced than they actually are.

The thing is - investors associate AI with “future growth”, consumers view AI features as '“cutting-edge”, and companies feel pressure to keep up with competitors adopting the label. Calling something “AI” signals “innovation”, even when the underlying system is little more than standard automation or basic machine learning.

As a result, the definition of AI has been modified a lot. Tools that would have once been described as algorithms or analytics systems are now rebranded under the AI umbrella.

There have already been public examples of this trend. A campaign by Coca-Cola drew criticism after claiming a new drink had been co-created with AI, without clearly explaining what role the technology actually played. Reporting by Forbes highlighted how the claim appeared to rely more on branding than substance.

Regulators are also beginning to respond. In 2024, the U.S. Securities and Exchange Commission charged two firms with making misleading statements about how extensively they used AI in their investment strategies.

The point is - AI has become a marketing label and like any label tied to hype, it gets stretched.

The Rise of Agentic AI

While much of the confusion around AI comes from marketing, the technology itself is still evolving. One of the latest developments is what many now call agentic AI.

Unlike traditional systems that wait for prompts, these models are designed to take action. They can coordinate and interact with different software tools to complete an objective:

• Generative AI models, such as tools like ChatGPT or image generators like Midjourney, focus on producing content.

• AI agents go a step further by using that intelligence to perform specific tasks, such as scheduling actions, retrieving data, or executing workflows across applications.

• Agentic AI extends this idea even further. It involves multiple agents working together, coordinating tasks and decisions with minimal human input.

Projects like OpenClaw highlight how multiple agents can coordinate to handle complex tasks. However, this space is still in its early stages. These systems often struggle with reliability and introduce serious security risks when given access to external environments.

Despite these limitations, progress is being made. Agentic systems are likely to improve and expand, making them an area worth watching closely.

Why Understanding This Matters

At a glance, the distinctions between different types of AI models and technologies can seem academic. In reality, they directly affect how AI systems behave.

A recommendation engine predicts preferences. A generative model produces content that may or may not be accurate. An agentic system can take action, which introduces a different level of risk and responsibility.

Knowing what kind of system you are dealing with, helps in setting expectations and answering questions like - can it make decisions on its own? Can it generate misleading outputs? Should its actions be verified?

It also helps filter out hype.

We also have terms like “singularity” - the idea that AI could surpass human intelligence and rapidly evolve beyond control - which are often used to create urgency and attention. In reality, current systems remain far from that scenario.

Understanding the layers of AI makes one thing clear - the technology is advancing, but it is still bounded. And recognizing those boundaries is what keeps expectations grounded.

What Comes Next

Much of today’s AI conversation revolves around how current systems are rapidly approaching Artificial General Intelligence (AGI).

Despite their capabilities, modern AI systems still struggle with core problems such as reasoning, long-term planning, and understanding cause & effect. They can generate convincing outputs, but that doesn’t mean they truly understand what they produce.

These limitations matter. They suggest that simply scaling current models may not automatically lead to human-level intelligence. And this doesn’t mean progress has stalled. It means expectations need to be calibrated.

In short, AI is not a single system or breakthrough. It is an umbrella term that covers a range of technologies built on top of each other - from machine learning to deep learning, from transformer architectures to generative models.

Each layer adds capability, but not all of them represent the same level of innovation.

Some, like generative models, have introduced visible and impactful changes in how people interact with technology. Others are slow improvements that refine existing systems rather than redefine them.

With that being said, thanks for reading.
Peace out.

One Last Thing

Nearly all people interact with systems without questioning them. This space is about doing the opposite aka understanding how things work, where they fail, and what that means in practice.

If that’s the kind of content you want more of, subscribe. And if this piece added something new to how you see things, pass it on. Good information only matters if it travels.

And if you’re active elsewhere, follow along on Bluesky or X (Twitter) or LinkedIn - those are the two places I show up most consistently outside Substack. A follow, a like, a repost - it matters to me, more than you think.

What You Might Like

• The Truth About the Dark Web: Myths vs Material

• How We Got Online: The Hidden History of the Internet

• What are LLMs & How Are They Changing Our World?

• Blockchain - The Cool Parts of the Tech Behind

Let’s begin with a question. What comes to mind when you hear of the term “gamer”?

My guess is, it’s someone on a couch with a controller in hands, TV glowing in the center room.

Or maybe it’s a PC setup with RGB lighting which alone costs more than my laptop.

That’s the general image gaming culture has spent decades building. But here’s the thing. That image is wrong. Or at least, it no longer holds true thanks to the rise of mobile gaming.

The biggest gaming platform in the world right now is sitting inside your pocket without many of us knowing.

It’s been there this whole time, we just weren’t paying that much attention. And it’s not just us, big players in the gaming industry ignored it for quite some time too.

Yet now, all the big players which includes Sony, Microsoft, Nintendo, and even Valve are all scrambling to get a piece of it.

Mobile gaming isn’t a side category anymore. It’s the main event. And most people have absolutely no idea.

Even though I wasn’t sure that the mobile gaming market is that big, when I saw the numbers for my recent article on Carcinization in the Gaming Industry, I was somewhat shocked.

Mobile gamers who most hardcore gamers don’t even like to call gamers have given rise to an industry that is double the revenue and 5x the size of the console market.

This realization was the idea for this article. Initially I thought of mentioning it in a small section in the Consoles and Carcinization article.

But there was a lot more to this topic so I had to dive deep and write a proper article explaining where this whole thing started and where the mobile gaming space is headed in the years to come.

The Surprising Numbers

I’ll just say it at the start. Mobile gaming made roughly $103 billion out of the total of $188 billion the entire gaming industry made in 2025. Yes, that’s BILLION with a B.

Just let that sink in for a second.

Console gaming which includes your PS5s, Xboxes and Nintendo Switches of the world pulled in $45 billion. And PC gaming came in 3rd pulling in $39 billion in 2025.

The entire global gaming market was worth roughly $188 billion in 2025, and mobile alone is carrying nearly half of that number. By itself. From the phones in people’s hands.

There are 3.4 billion mobile gamers in the world right now. Compare that to roughly 630 million console players. Mobile gaming isn’t just bigger, it’s in a completely different league.

And yet, if you asked most people around you to name the biggest gaming platform, I’d bet they’d say PlayStation or Xbox before they even thought about their phone. That gap between what we think and what the numbers actually are, is exactly what this article is about.

Isn’t Mobile Gaming just Candy Crush?

I know what you’re thinking. Okay, a lot of people play phone games. But it’s all Candy Crush and Subway Surfers and ads every 30 seconds. That’s not real gaming.

I used to think the same thing, honestly. I would laugh at mobile guys even calling themselves “gamers”. But the numbers and the impact speak for themselves.

It’s hard to find students who in the COVID days haven’t tried PUBG Mobile at least once. Even if you didn’t play it you almost certainly have heard it make headlines.

I remember jumping ranks and grinding for new costumes like it was my full-time job during lockdown days :p

PUBG Mobile generated over $3 billion in revenue in a single year. Mobile Legends has tournaments with prize pools that rival traditional Esports.

These are not simple games, not even cheap to make actually. They’re built by massive studios with hundreds of developers, cinematic trailers, and dedicated competitive scenes even in countries like Pakistan.

Honor of Kings, a mobile MOBA (Multiplayer Online Battle Arena) also made by Tencent crossed $2 billion in annual revenue in 2024 with over 100 million daily players.

Just for context, that’s One Hundred Million people playing one game, every single day. That’s not casual. That’s a media empire that basically lives on our phones.

Genshin Impact crossed $10 billion in lifetime player spending faster than any mobile game in history.

And mind you, this game has console-quality graphics, a massive open world, regular content updates, and a player base that would make some AAA studios jealous.

The Candy Crush phenomenon kinda exists in the sense that casual games loaded with ads still exist and make a lot of money. But dismissing actual mobile gaming because of these games is just not a good strategy.

The Math Is Mathing

Let’s actually break down where mobile’s $92.6 billion in the year 2024 comes from because the breakdown seriously hits the eyeballs.

In-app purchases made up about $80.9 billion of that $92.6 billion number. Advertising brought in $18.9 billion. Subscriptions are a small but growing slice on top of that.

Here’s the interesting part though. Downloads are actually going down. There were 49 billion mobile game downloads in 2024, which is down 7% from the year before.

Fewer people are downloading games yet the spending is going up, sounds weird, right?.

But that just means that existing players are spending more money per person, not that new people are flooding in. And that’s what makes the total number go up.

Strategy games led all mobile genres at $17.5 billion in revenue. This includes games like Clash of Clans or the one I used to play, Clash Royale.

RPGs followed strategy games at $16.8 billion. These are not Candy Crush numbers. These are genres that require real time investment, learning, and often a decent amount of money spent inside the game.

And this isn’t slowing down anytime soon. Mobile gaming is projected to go over $130 billion in 2026 and the numbers only balloon up when you look at projections for the end of this decade.

If 2024 shows roughly $90 billion revenue, just imagine what the numbers would look like by 2030 if the mobile gaming industry keeps on growing as is.

So, Who Actually Plays Mobile Games?

Here’s where it gets really interesting from a global perspective.

Asia Pacific (which includes our Pakistan) generated 49% of all mobile gaming revenue in 2025.

Basically, nearly half the world’s mobile gaming money comes from one region. China alone generates over $40 billion annually, which is more than the entire console market in most other countries.

India is the next one to watch. 450 million gamers, getting by with their sub-$100 smartphones and some of the cheapest mobile data plans in the world.

Despite that massive player base though, India generates only 1.1% of global gaming revenue right now.

That gap between the number of players and the amount of money being spent tells you exactly where the next wave of growth is coming from. Game publishers are already paying very close attention.

Japan and South Korea punch way above their weight. Average revenue per user in those markets is above $80, which more than makes up for their smaller populations.

But here’s the stat I find most interesting whenever this topic comes up. The average mobile gamer is 36 years old. And 46% of gamers worldwide are women.

The image of a skinny teenager in a dark room with a headset was simply never accurate for mobile.

The Business Model that Prints Money

Okay so we know mobile gaming is massive. But how exactly is it making this much money? Because most of these games are free to download. I’ve played games like Clash Royale and I’ve never spent a single rupee.

The answer is the free-to-play model, and once you understand how it works you’ll start seeing it everywhere.

Zero upfront cost removes the biggest barrier to trying anything. You don’t have to decide if a game is worth Rs. 3,000 before you’ve played it. You just download it and start playing.

The risk of trying a game is zero and so are the requirements when it comes to your smartphone. Most mobile games run well on phones that aren’t even bleeding-edge.

Once you’re playing and invested in the game - your characters, your progress, your ranking - that’s when the real product reveals itself. Skins, extra lives, characters, battle passes, random loot drops you pay for.

That’s what got me hooked on PUBG in lockdown days. Once you’ve spent time pas a couple seasons and your character has leveled up, you just want to keep going and spending real money enters the equation sooner or later.

This is the reason In-app purchases brought in $80.9 billion to the mobile gaming market in 2024 alone.

The gacha mechanic deserves a special mention because it’s basically a slot machine dressed up as a game feature.

If you haven’t heard of the gacha system or want to know more about how gambling is implemented in video games, check out this recent article from my fellow writer Mohib here at SK NEXUS.

In short, you spend currency - either earned slowly in-game or bought with real money for a random chance at a rare item. Basically, how lootboxes work in Counter-Strike.

Genshin Impact built its entire economy around this. Players have spent thousands of dollars chasing specific characters.

One player famously spent over $10,000 on a single banner without getting the character they wanted. Now, that’s a shame :/

However the regulators are starting to notice this. The FTC hit Epic Games with a $520 million settlement over Fortnite’s use of dark patterns - design choices that tricked players, including kids, into spending money they didn’t mean to spend.

Valve also just got hit with a lawsuit over Counter-Strike lootboxes. Europe is actively debating loot box regulation.

This is a business model that has worked for quite some time and still works brilliantly, but it’s starting to face its first real legal challenges.

Sony is waking up to mobile, slowly and awkwardly

Sony is one of the most interesting cases in this whole story because they clearly know what they need to do, they’re just not very comfortable doing it.

In 2022, Sony set up a dedicated PlayStation Mobile division and acquired a studio called Savage Game Studios specifically to build mobile titles.

Their internal target is to have 20% of PlayStation’s first-party output on mobile by 2026. That’s a very specific number for something companies only announce when they’re serious.

Although Sony’s problem isn’t money or talent. It’s culture.

PlayStation built its identity on prestige console experiences. God of War, Spider-Man, The Last of Us.

The internal feeling at many of these studios is that mobile is a step down. Getting developers who spent years creating 40-hour story-driven games to suddenly care about a free-to-play mobile experience is a genuine challenge.

But Sony watched Call of Duty Mobile and Diablo Immortal - both are mobile versions of massive console franchises and they become billion-dollar mobile games, that’s hard to ignore for Sony.

Sony has an incredible IP sitting on the shelf. God of War on mobile is not an insane idea. It’s actually a very logical one if the implementation is right.

Sony has marked roughly $12 billion for strategic investments between 2024 and 2026. Mobile is explicitly mentioned on that list.

Microsoft Wants Xbox Everywhere, Literally!

Microsoft’s strategy is more aggressive and more interesting than Sony’s, partly because they already made their big move.

The Activision Blizzard acquisition for $68.7 billion (the largest acquisition in gaming history) was not primarily about console gaming. Microsoft paid that much partly to get King.

King is the studio behind Candy Crush, and Candy Crush alone generates over $2 billion annually just from mobile. Microsoft bought an entire mobile gaming business as part of that deal.

Beyond that, Microsoft has been building toward a world where “Xbox” is not a box you plug into your TV but a service you access from anywhere.

Xbox Cloud Gaming lets you stream games to your phone right now. The quality is genuinely good if you have a stable connection. Sadly, we in this part of the world shouldn’t be too happy about cloud gaming anytime soon.

Microsoft tried to build a dedicated Xbox mobile store for Android but hit a wall because Google’s Play Store policies blocked it. That legal battle is still going on.

If the legal battle resolves in Microsoft’s favor, they could sell Xbox games directly on Android phones without giving Google a cut. That is a very very big deal.

Microsoft’s pitch is simple - you own a game, you should be able to play it anywhere. On your console, on your PC, on your phone.

That actually sounds cool on the surface and mobile is central to making it work. But I won’t be getting too hyped about Microsoft anytime soon I guess.

Nintendo - The Silent Player

People talk about Sony and Microsoft’s mobile strategies like they’re new ideas. Nintendo figured this out almost a decade ago and most people missed it.

Mario Run launched in 2016. Fire Emblem Heroes launched in 2017 and has quietly generated over $1 billion in revenue since then.

Almost nobody talks about that number. Animal Crossing: Pocket Camp built a dedicated mobile audience that existed completely separately from the console version.

But Nintendo’s mobile strategy is genuinely different from Sony and Microsoft’s.

Nintendo doesn’t use mobile to replace console revenue. They use it as a marketing funnel to sell more Nintendo Switches.

Their strategy is to get someone hooked on a mobile Fire Emblem game and they’re a much easier sell for the full console experience.

The Switch itself is maybe the clearest expression of Nintendo’s philosophy.

It basically is a console you can take anywhere, play on your TV at home, or hold in your hands on a bus. It blurs the line between console and mobile gaming in a way that nothing else does, maybe except the SteamDeck.

The Switch 2 sold 1.6 million units in its first month in the US alone in 2025, which tells you the hybrid model is very much working.

Nintendo is not chasing mobile the way Sony and Microsoft are. They already built their mobile strategy into the hardware.

Valve Just Joined The Chat

Valve dropped a bombshell when it recently announced its upcoming hardware. People have been onto the hardware ever since, but there’s a fresh news that most people haven’t fully processed yet.

In November 2025, Valve announced the Steam Frame - their upcoming VR headset running on a Snapdragon ARM chip that natively supports Android APKs.

The Steam Frame is a different beast and pretty much the most exciting thing that’s going to happen to gaming on VR. I wrote an in-depth article discussing its features linked here.

What supporting Android APK means in plain language is that Android mobile games can run directly on Steam without any porting or conversion work from developers.

A Valve engineer (Jeremy Selan) described the goal as making it so players “don’t even have to think about” what device their game is running on. Focusing on just that sentence tells you Valve’s entire strategy.

Valve has also been quietly funding Fex, an open-source Windows emulator for Android. This powers something called GameHub, which lets you play Steam games on your phone.

It’s not a finished product yet and may not launch with the Steam Frame at launch, but Valve confirmed that it’s serious about it. And considering the possible delays of Valve hardware, it’s possible that this may just launch with the hardware (if it releases)

The long play here is potentially quite massive for Valve and gamers in-general.

If Google’s Play Store faces enough regulatory pressure to open up (which is very much being debated in courts right now) Steam could start selling games directly on Android phones.

With no cut to give to Google that would mean Steam’s existing library available on your phone, bought through Steam themselves.

If this happens this would completely change the mobile gaming landscape.

Pakistan and Mobile Gaming

There’s a reason Asia is the leader in mobile gaming adoption. Countries including Pakistan, India, Bangladesh have large smartphone adoption and low-incomes to justify expensive PCs and consoles for video games alone.

This is part of the reason why Pakistan has over 34 million online gamers, 60% percent of which are mobile gamers while the 30% are on PCs and the rest of the 10% on gaming consoles.

As we’ve discussed many times before, Pakistanis mostly earn in PKR and spend in USD so the majority of households are kicked out of owning dedicated gaming machines.

But smartphone adoption is already at roughly 70-80% of the entire 250 million population of Pakistan. And that’s what makes our local mobile gaming scene more exciting.

Fortunately, we’ve also seen an increase in gaming overall through world champions like Arslan Ash and that has allowed for a surge in gaming in-general in Pakistan.

This has also translated to mobile gaming with big banks like HBL regularly sponsoring gaming tournaments as well international platforms like IGN also being available here.

Pakistan has also started developing a very small game development scene that also covers mobile games but that space is quite niche and is quite recent.

That’s not to say all is hunky and dory. Some game creators have found it difficult to monetize their work especially because of the low in-game ads that pay out 1-1.3 cents for a 1,000 impressions compared to $20 in USA and similar developed markets.

Even though we have yet to produce a mobile game that makes it big we still can be happy about some progress being made in this area.

The fact that we have started to look towards promoting any digital industry in the background of brick and mortar industries like textiles and clothing.

The Dark Side to Mobile Gaming

I’ve spent this whole article making mobile gaming sound like the greatest thing since sliced bread. And look, the numbers are real, the growth is too.

But I’d be lying if I wrapped this up without talking about what’s happening behind the scenes.

Because mobile gaming, more than any other gaming platform, has a serious dark side. And it’s one that the industry has been very deliberately quiet about up till pretty recently.

Let’s start with how these games are actually designed.

The term you need to know is “dark patterns.” It sounds like something from a sci-fi movie but it basically means design choices that are intentionally built to confuse or trick you into spending money you didn’t mean to spend.

In mobile games this shows up in sneaky ways. Buttons placed so close together that you accidentally tap the “buy” option when you meant to close a menu.

Games that wake from sleep mode and charge you before you’ve even looked at the screen. Currencies that are deliberately confusing - you buy 1,200 “gems” for Rs. 1,500, and then a character costs 950 gems, so you can never quite figure out what you’re actually spending in real money.

This is not a conspiracy theory. The FTC sued Epic Games over this exact behavior and won.

Epic paid $245 million in consumer refunds and a separate $275 million penalty for collecting data from kids under 13 without telling their parents - a total of $520 million, the largest gaming settlement in FTC history.

The complaint described how Fortnite’s button layout led players to get charged while “attempting to wake the game from sleep mode, while the game was in a loading screen, or by pressing an adjacent button.”

And then there’s the kids problem. Mobile gaming is everywhere. The phone is the most accessible screen in any household. A seven-year-old doesn’t need to ask their parents to switch on their iPad - they just pick up their tablet sitting on the table and open a game.

Most of these games are designed for all ages on the surface, but the monetization underneath is built for adults with credit cards. The problem is that kids don’t always know whose credit card is whose.

There are parents who have discovered hundreds of thousands of rupees worth of in-app purchases on their bills, made by children who genuinely didn’t understand they were spending real money.

The virtual currency systems in these games are specifically designed to create that confusion. You’re not spending money, you’re spending “coins” or “gems” or “V-Bucks.” It feels like Monopoly money until the bank statement arrives.

The addiction angle is also something the research community has been taking increasingly seriously.

The World Health Organization formally included Gaming Disorder in the International Classification of Diseases in 2018, recognizing it as a legitimate condition and not just parents being dramatic.

Research specifically on mobile gaming found that long hours of phone gaming in particular (more than computer or console gaming) showed the strongest links to poor well-being scores in adolescents.

The phone is always there. There’s no natural stopping point the way there is when you turn off a TV. You’re lying in bed at 2 AM and the game is right there in your hand.

And honestly, I’m a witness to this. My 4 year-old nephew has his eyeballs glued to his tablet playing Talking Tom Gold Run. And especially while playing it, it’s like he’s just hypnotized into the game.

China, which generates more mobile gaming revenue than almost anywhere else, got so worried about this that the government stepped in.

They introduced hard limits on how long minors can play online games - a maximum of one hour on weekdays and three hours on weekends, with a complete ban between 10PM and 8AM.

That’s the government of the world’s biggest mobile gaming market deciding the industry couldn’t regulate itself.

Europe is actively working on loot box regulation. Belgium already banned them outright in 2018. The Netherlands followed. More countries are in the conversation.

None of this means mobile gaming is bad. It means it’s a powerful thing, and powerful things need honest conversations around them.

The same way we talk about social media’s effect on teenagers, or the way junk food is marketed to kids, mobile gaming deserves the same scrutiny.

The business model works brilliantly for the companies. The question is who bears the cost when it works too well.

The industry knows all of this. They have researchers and psychologists on staff whose literal job is to figure out how to keep you playing longer and spending more.

That’s not speculation - it also came out in the Epic lawsuit. Internal documents showed employees flagging the dark patterns as problems, and being ignored for years.

That’s the part of the $92 billion story that usually doesn’t make it into the press releases. But progress is being made on this end too.

What The Future Looks Like?

Moving back to the global landscape, here’s where we currently are:

Mobile gaming is already the biggest platform, both by the revenue and the number of players. The money in mobile gaming speaks for itself.

What’s changing now is that the companies who spent decades building consoles and PC gaming are finally accepting that reality and building towards it.

Microsoft bought King.
Sony built a mobile division.
Nintendo made the Switch.
Valve is building Android support into Steam.

These are not coincidences. These are companies that read the same revenue numbers you just read, and decided they needed to be in that market. This is a battle to establish foot into the mobile market.

Cloud gaming is the bridge that makes all of this work at scale. Xbox Cloud Gaming, PlayStation Remote Play, NVIDIA GeForce Now - these services remove the hardware requirement entirely.

You don’t need a console. You don’t need a gaming PC. You need a phone and a decent internet connection and even though Cloud gaming sucked initially, it’s gotten way better.

5G is making that connection reliable enough for real-time gaming in more and more places. South Korea and China already have gaming-grade 5G network slices with sub-20ms latency.

This is the same latency we Pakistanis get on our broadband connections at our homes.

As that infrastructure spreads, the gap between mobile gaming and console gaming shrinks further. And in that regard we’re also hearing about the 5G auction in Pakistan although I don’t really know how it materializes.

But still, good to know I guess :p

And then there’s Africa, which is projected to grow at 12.5% annually through 2031 in mobile gaming.

The next billion players are coming from markets that are going to skip consoles entirely - the same way large parts of the world skipped landlines and went straight to mobile phones.

The lines between console, PC, and mobile are genuinely starting to disappear. Valve’s Steam Frame is maybe the clearest single example of that. But it’s happening everywhere, slowly and then all at once.

Mobile gaming didn’t sneak up on the gaming industry. It looks like the industry was just too self-involved and now mobile is out there giving everyone reality checks.

As always Thanks for taking the time to go through this deep-dive article. Every month we do one of this going head-first into topics we find interesting.

I love reading your comments so feel free to leave your feedback down below.

And before you leave here’s a question for you:

Further Reading and Resources

• Sony sets up PlayStation mobile gaming unit

• Naavik - PlayStation strategy deep dive

• Valve is welcoming Android games into Steam

• Gaming – Pakistan’s New Digital Entertainment Playground - DAWN

• Unlocking Pakistan’s Digital Future

• Newzoo’s Global Games Market Report 2024

• Statista mobile gaming revenue breakdown

Remember when software used to just…work?

You would install something, learn how it works once, and it would keep doing exactly what you expected. There were no constant redesigns or algorithms secretly deciding what you should or shouldn’t see.

But, those days seem to be in the past now because today - most platforms feel harder to use despite being more ‘advanced.’ Search results feel polluted. Feeds are filled with things you didn’t ask for and interfaces change for no clear reason.

You must have felt this as well…there’s this strange situation where everything works, but the experience keeps getting worse. NOT surprisingly, there’s a term for it - Ensh*ttification.

What Ensh*ttification Actually Means

The term Ensh*ttification, popularized by Cory Doctorow, describes the gradual decline of digital platforms as they shift their priorities away from users and toward profit extraction.

At a surface level, ist sounds like a rough way of saying ‘things are getting worse.’ But the idea goes deeper than that. It describes a predictable lifecycle - one where platforms follow a structured path as incentives change.

Before going further, there’s an important distinction to make.

A lot of people confuse Ensh*ttification with planned obsolescence.

Planned obsolescence is about hardware. Devices slow down or wear out over time. While Ensh*ttification is about software. In one case, the product physically degrades. In the other, the product continues to function - but increasingly works against the user.

To understand why that happens, you have to look at how platforms evolve over time.

Stage 1 - User-First

In the beginning, platforms are built to attract users.

Different platforms focus on building something that people want to use. The experience feels clean & valuable. The features are designed to solve real problems and most importantly the growth comes from usability, not manipulation.

This is where trust is built, but it’s also the bait.

At this stage, companies market their product in a way that it feels promising, feels useful, something that respects your time. And for a while, that promise holds.

Stage 2 - Ecosystem Shift

Once a platform has enough users, the priorities begin to change.

This is where the ecosystem starts changing. Businesses, advertisers, and creators become the new focus because they generate revenue. The platform begins to reshape itself around them.

You start seeing algorithmic control replacing organic reach. Content is no longer shown because it’s relevant, but because it serves a business goal. Pay-to-play systems emerge, where visibility increasingly depends on spending money.

The change is gradual, which makes it harder to notice all this happening.

Stage 3 - Extraction Mode

Eventually, the platform runs out of easy growth.

At this point, it turns inward.

Even business customers start getting squeezed. Ad loads and costs start to shoot up. Reach declines unless you pay more. Features are redesigned to maximize engagement.

This is where the platform reveals its true purpose which is to extract as much value as possible from everyone.

Why This is a Business Model - Not a Bug

Network Effects

Every major platform is powered by other people.

The reason a platform becomes valuable is because of who is already using it. If your friends are there, you join. If your audience is there, you stay. If your customers depend on it, you build around it.

With the passage of time, this compounds. The platform becomes the place where interactions happen by default and that exact concentration gives the platform leverage. On top of that, leaving stops being a simple choice. Even if a better alternative exists, it feels empty without the same network.

Switching Costs

Once people are in, switching becomes difficult in ways that are not always obvious.

There is the visible layer, that involves your data, messages…etc. But the heavier layer is everything built on top of that aka your identity, your presence, your connections.

And it’s really hard to move that cleanly.

Exporting data rarely recreates the experience somewhere else. A list of followers is not the same as an active audience. A backup of messages does not replace ongoing conversations and that’s why leaving starts to feel like starting over and who likes that?

That friction is enough to keep most people where they are, even when the experience declines.

Monopoly Pressure

The environment around the platforms start to change once they grow. In the early stages, there is pressure to improve. Users can leave easily, so companies focus on making the product better and more useful.

Over time, that pressure weakens. New players struggle to attract users because the network is already established elsewhere. The platform settles into a position where users stay stuck in the loop because alternatives feel impractical.

At that point, priorities relocate internally. Decisions are no longer shaped by user experience in the same way. Control becomes easier to enforce and changes can be pushed without the same risk of users walking away.

Shareholder Growth

The final pressure comes from expectations around growth.

Large platforms are not designed to stay stable. They are expected to increase revenue consistently. In the early phase, that growth comes from new users and expanding markets.

But that phase does not last forever and there is a limit to how many people can join and how much time they can spend. Once that limit is reached, growth has to come from somewhere else.

In order to solve this issue, the platform is made sh****er - more ads are introduced into the same space. Features that were once freely available become restricted or monetized. Visibility becomes something that can be bought.

None of these changes happen all at once. They are introduced gradually, often framed as improvements.

From the outside, the experience feels heavier and less intuitive. From the inside, the system is doing what it was built to do aka extract more value from the same users.

Putting It Together

None of these forces act alone. They stack on top of each other, and they unfold in a sequence.

A platform starts by being useful. That’s how it pulls people in. Once enough users are in, network effects and switching costs begin to take hold. Leaving becomes inconvenient, then difficult, and eventually unrealistic for most people.

As that grip strengthens, competition becomes less relevant. Now, the platform doesn’t need to win users the same way anymore because it already has them, due to which the strategy relocates toward pulling more value out of the same system.

And as a result what you get is a product that still works, but it feels worse.

You can see how widely recognized this pattern has become. Even the Government of Norway acknowledged it with a video. It’s funny, but it lands because it reflects how these systems actually behave. You can watch it here:

Case Studies

The easiest way to understand Ensh*ttification is to watch it play out in systems you already use. The pattern is really visible and in many cases, hard to avoid:

Facebook (now Meta)

Meta (previously Facebook) started as something which wasn’t available to the general public. But when that changed, early users were able to connect with people they cared about, and see what they shared.

That phase did one thing very well. It attracted users and locked them in.

As more people joined, the platform became harder to leave. Everyone’s friends, history, communication channels - all were at one place. Leaving meant disconnecting from all of it at once and nobody wanted that.

Once the users were locked-in, the priorities changed.

Meta began reallocating attention. The feed stopped being a direct line to people you followed and turned into a system optimized for engagement and monetization. Ads became more frequent. Posts from publishers and brands were inserted into spaces users didn’t explicitly choose. Behind the scenes, the platform expanded its data collection to make that targeting more precise

Publishers and advertisers came in next. Meta had the audience, and that was enough. Businesses followed because that’s where attention was. Media companies started relying on it for traffic.

Then the algorithm took control.

Instead of showing you posts from people you chose to follow, the system started deciding what you should see. What keeps you scrolling gets pushed. What doesn’t, disappears.

You can already see the next version of this on Instagram. The feeds are fully curated, ads are baked in and what you see is no longer a reflection of who you follow, but what the system wants you to engage with.

Even core ideas like privacy are starting to change. Encryption, which was once pushed as a feature, is now something platforms are under pressure to weaken or remove entirely.

Amazon

Amazon followed a similar path, just in a different domain.

In its early years, Amazon focused on being the most convenient place to shop. Prices were aggressively low, often below cost. Shipping was fast. Search worked the way users expected (relevant products showed up first)

That combination pulled users in and changed behavior. For many, Amazon became the default starting point for shopping.

Lock-in came through habit and ecosystem. Services like Prime encouraged users to stay within the platform. Online purchases added another layer. Books, audiobooks, and other media were tied to Amazon’s systems; if users wanted to leave amazon, they would have to give up everything, aka making it costly to leave.

Once the user base stabilized, suppliers became the next focus.

Third-party sellers joined to access that demand. Amazon offered reach, logistics, and infrastructure. For a while, the terms were favorable enough to make the platform attractive. Sellers scaled their businesses around it, but things didn’t stay like that forever.

Fees increased, visibility inside search results became tied to ad spend. Sellers started paying to be visible on the platform. In many categories, the top results are now dominated by sponsored listings rather than the most relevant products.

Now, we have a platform on which users still find what they need. Sellers still make sales. But both operate inside a system that steadily extracts more value from them than it returns.

Further Examples

This pattern isn’t limited to social media or e-commerce. You see it across almost every digital service.

Take streaming platforms. Netflix, Hulu, and others followed the same. They started by offering convenience and large content libraries at reasonable prices. Over time, prices increased and access became conditional. What you “have” depends on what you’re currently “paying for.”

Ownership, pretty much, disappeared in the background. Because think about it - the moment you stop paying, it’s gone. The same logic applies to software, games, and even devices.

Search engines show a similar shift. Look at Google. A simple query used to return the most relevant results. Now, the top layer is often sponsored placements or SEO-optimized pages competing for visibility.

Then there’s the ecosystem effect.

Every platform out there is on a mission to build a full network of services around you. If you use Gmail, you’re likely using Google Photos, Google Drive, maybe even Android. Each product reinforces the other. Leaving one means disrupting all of them.

You see a similar dynamic outside of software.

Take modern cars. Features that used to be part of the product are now tied to subscriptions. Different features are now being locked behind ongoing payments. You buy the car, but access to parts of it depends on whether you keep paying.

Companies understand this. Get users in, build dependence, then change the terms.

Where this is Heading

I can’t see into the future, but from where I sit, it seems - monetization will get more aggressive. Systems will extract more from the same user base rather than expanding it. Algorithms will become harder to understand, making it difficult to know why you’re seeing what you see. Creators will become more dependent on platforms they don’t control. Reach, visibility, and income will stay tied to systems that can change the rules at any time.

AI will accelerate all of this. Feeds will be filled with synthetic content optimized for engagement.

And here’s another interesting thing I want to bring to your attention:

In February 2026, the United States Congress moved forward with requirements tied to impaired driving prevention systems in future vehicles. These systems are designed to monitor drivers and intervene when needed. The goal is “safety”, but the mechanism is constant monitoring built into the product itself.

Everything I mentioned here, isn’t directly tied to Ensh*ttification, but it points in a similar direction - Control is moving away from the user and everything is getting worse.

Across platforms, devices, and services, the model is becoming consistent and given how these systems have evolved so far, there’s little reason to expect that to reverse anytime soon.

What You Can Do About This

Platforms control the environment, but you still control what you feed into it. Don’t put all your eggs in one basket. Relying on a single platform for audience, communication, or content locks you in. Spread yourself across multiple tools and services so leaving one doesn’t mean losing everything.

Own what you can. If you’re building an audience or creating content, keep a layer that’s yours. Email lists, local backups, personal storage…etc Anything outside the platform’s walls gives you leverage. The less you depend entirely on someone else’s system, the more freedom you keep.

Support alternatives when possible. Smaller platforms, open-source tools, or services that don’t monetize through extraction are worth exploring. Doing this might not be convenient, but know that even convenience comes at the cost of control and security being taken away.

Finally, pay attention to how your decisions are shaped. Algorithms nudge your behavior. Be skeptical of what you see because most of what you see is designed to make you act without thinking. Recognizing this lets you step back and act deliberately instead of reacting.

Platforms will keep evolving. They’ll attract, lock in, extract. Understanding the cycle doesn’t fix it, but it gives you a way to operate with intention rather than being dragged along. That’s how you keep some control in a system designed to take it.

Thanks for reading.
See you next time.

One Last Thing

Nearly all people interact with systems without questioning them. This space is about doing the opposite aka understanding how things work, where they fail, and what that means in practice.

If that’s the kind of content you want more of, subscribe. And if this piece added something new to how you see things, pass it on. Good information only matters if it travels.

And if you’re active elsewhere, follow along on Bluesky or X (Twitter) or LinkedIn - those are the two places I show up most consistently outside Substack. A follow, a like, a repost - it matters to me, more than you think.

What You Might Like

• Privacy Is Not Security - The Difference is Important

• Gambling With Better PR - Prediction Markets

• How Gambling Went Digital

• The Business of Trust - Influencer Marketing

Let me begin with a question:
How many times have you picked up your phone today?
Ten times? Twenty? More?
Do you ever feel like your battery is Goodenough?

Every single one of those times, a tiny chemical reaction fired up inside your device to give it the power to work. You did not think about it. I don’t think about it and maybe, no-one else does.

And that is kind of the point of this article today.

Batteries are some of the most invisible pieces of technology in most of our tech products. And they are also some of the most important ones especially in mobile devices.

Everything from smartphones, laptops, to even our BP monitors run on batteries. Especially in tech, the battery dictates our purchase decisions too.

The thing is that we are at a genuinely interesting moment in battery history right now because the chemistry that has powered your phone, laptop, and earbuds for the last thirty years is hitting its limits.

There is an innovation race going on between China and the US, as one controls the entire supply chain of batteries we use today while the other is working hard on challenging that with a new innovation.

And in that race we’re seeing some of the most interesting developments when it comes to the technology that will power our devices going forward.

So that’s what I’ll talk about in today’s article.

What a Battery Actually Is?

Before getting into the cool stuff let’s try to see what a battery actually is.

To explain it simply, a battery is just a controlled chemical reaction that produces electricity.

That is it. That is the whole thing. Not really but you get the point.

Actually, two metal points - a positive one called the cathode and a negative one called the anode - sit inside a material called an electrolyte.

Ions (charged atoms basically) move between the two electrodes through that electrolyte, and that movement creates an electric current that powers your device.

Rechargeable batteries work by reversing that reaction when you plug in. The ions travel back to where they started, and the battery is ready to go again.

Non-rechargeable batteries like your AA Toshiba cells that you get from your local hardware shop cannot reverse this reaction.

Once the chemistry is done, it is done. They are disposable and that is why you throw them away.

The whole field of battery technology is basically about finding materials that can do this reaction better - store more energy, do it more safely, and do it more times before wearing out.

Just a Pinch of History

I know that most of you think of history as a boring subject and I won’t bore you guys much but just bear with me for a second.

So we know that the first real battery was built by an Italian scientist named Alessandro Volta in 1800.

Volts were named after our guy Volta.

So Volta basically stacked copper and zinc discs on top of each other, separated by cloth soaked in saltwater. That was the first battery as we know them.

It produced a steady electric current and the scientific community lost its mind because nothing like that had ever existed.

Volta called his invention ‘The voltaic pile’ It was clunky. It leaked. It corroded quickly. But it worked. And off course it took the world by storm.

From there however, the progress was slow but steady.

The invention of battery was a disruption and after that progress started to take its course inside battery technology.

In 1859, a French physicist named Gaston Plante invented the lead-acid battery - the first rechargeable one.

The same battery design that Plante invented still sits in your car right now, 165 years later. That tells you how good it was.

Around 1899 we had another guy named Waldemar Jungner who invented the nickel-cadmium battery. These batteries were lighter than lead-acid, rechargeable, and eventually banned in many places because cadmium is extremely toxic :/

We still find Nickel-cadmium (NiCd) batteries as smaller rechargeable cells or in toys, digital cameras, and flashlights, etc.

And then almost nothing major happened in battery tech for nearly a century.

Progress happened slowly. Electronics kept evolving.

Batteries mostly stayed the same until Lithium-Ion entered the chat. From a chemistry POV, Lithium-ion batteries were fundamentally different from their predecessors.

Sony was the first company to commercialize the lithium-ion technology back in 1991 and this was the turning point for the industry and everything for battery tech ever-since.

Why Lithium-ion Specifically Won

Lithium-Ion batteries were a totally new chemistry back when they started to be commercialized in the early 90s.

Lithium-Ion batteries are named so because of their use of lithium and also because of lithium being the lightest metal on the periodic table.

That matters because when you are building a battery that needs to sit inside a phone or a laptop, weight is everything. Who appreciates a heavy phone or a laptop anyways?

Lighter material means more energy stored per gram. This property is known as energy density and lithium-ion’s energy density was unlike any battery that had come before it.

Sony’s first commercial lithium-ion cell in 1991 had an energy density of around 80 Wh/kg. That sounds like nothing but it was a revelation compared to what existed before.

I can go about explaining the unit of energy density but since it’s not a math article I won’t bore you with the details. For now you can just focus on the digit before the unit to get a general idea of how batteries improve over time.

Today’s best lithium-ion cells are above 300 Wh/kg. Almost four times better. From the same basic chemistry (lithium-ion), just refined over thirty years of engineering.

That refinement is what gave us better smartphones. Laptops that last more than ten hours. Wireless earbuds. Electric vehicles.

All of it runs on the same fundamental idea Volta had when he stacked those coins, just made dramatically better by lithium.

There is one catch though.

Lithium-ion uses a liquid electrolyte inside. And the problem with that liquid is that it can be flammable. That is why phones occasionally catch fire.

This is exactly why airlines have specific rules about lithium batteries in checked luggage. That is why every electric car fire you have seen in the news burned so intensely.

Lithium ion is a brilliant technology but it’s dangerous because that liquid electrolyte inside it can burn and does burn.

Still though everything from our Macbooks, iPhones, Solar systems use Lithium-Ion batteries now. It’s the de-facto choice of our time.

The Big Problem with Lithium Ion

Here is where things get uncomfortable for lithium-ion chemistry.

We have been working with the same fundamental battery chemistry for over thirty years now, that shows that it works.

And the graphite anodes inside most lithium-ion batteries today, graphite being the material that stores the lithium ions on the negative side, are approaching their theoretical max basically.

You cannot squeeze much more out of graphite. Physics will not allow it. And then there is the supply chain problem.

Most of the world’s lithium comes from Chile, Australia, and China. The mining process is water-intensive, and a lot of it happens in some of the driest places on earth.

Cobalt, another key ingredient in lithium-ion batteries, is mostly mined in Congo under conditions that have attracted serious international scrutiny for years.

None of this is a secret. The battery industry knows about it. Governments know about it.

It doesn’t matter when companies like Apple say that they value the climate, or the conditions of miners who source them the raw material for batteries in their products. That’s just how things are at the moment.

And then consider the scale of what is coming. Electric vehicles need batteries. Energy storage solutions like your WAPDA or even your solar setup at home need batteries.

By the way, If you’re someone like me who’s tired of WAPDA burning your tech and want to know how to protect from WAPDA killing your electronics, you can checkout a recent article I wrote here at SK NEXUS.

Every new smartphone, AI gadget, basically every new product category in tech needs batteries.

Lithium-ion in its current form cannot scale to meet all of that demand. Not cleanly, not safely, not at the pace we need it to be.

Silicon-carbon Batteries Are Already Here

We know that we’re hitting the physics of lithium-ion batteries. And this was the moment where many of the smart people knew we had to build something new.

The first real upgrade to lithium-ion in decades is not some exotic new chemistry from a research lab. It is a smarter material choice.

Standard lithium-ion batteries use graphite as the anode material. Silicon can store about ten times more lithium ions than graphite. That means more energy in the same space, which means a bigger battery inside the same size phone.

The problem with pure silicon is that it swells up to three times its own size during charging and then shrinks back during discharge.

Do that a few hundred times and the material cracks, the battery degrades fast, and you end up with a phone that barely holds a charge after six months.

The fix researchers landed on is clever. Blend silicon with carbon in a new structure. The carbon buffers the expansion, keeps the silicon stable, and retains most of the energy storage benefit.

That is what silicon-carbon anodes are. That is what is quietly showing up in your phones right now.

Samsung is reportedly targeting silicon-carbon batteries for the Galaxy S27 in 2027.

The Redmi Note 14 Pro’s silicon-carbon battery is 24% larger in capacity than its predecessor while physically occupying the same space.

The OPPO Find N5 used silicon-carbon to fit a larger battery inside a foldable that is just 4.21mm thin when unfolded.

Silicon-carbon batteries pack around 40 to 55% more energy in the same volume compared to standard graphite lithium-ion.

The battery tech in your next phone will not just be marginally better. It will be noticeably, meaningfully better. Without the phone getting any thicker.

You might not realize it that much because for an average person smartphones are already at the point where average phones easily last a day.

But laptops especially Windows laptops could greatly benefit from silicon-carbon batteries that store more while weighing around the same as previous ones.

The thing about silicon carbon batteries is that it’s not just in the future. It is happening right now. We have small smartphone players testing these with their products while big ones like Apple and Samsung wait it out for the results to mature.

Solid-state Batteries Are Kinda Here Too

Silicon-carbon batteries are an evolution of lithium-ion. But Solid-state is a different category entirely.

Right now, every battery you own has a liquid electrolyte inside - the material that lets ions move between the two electrodes.

That liquid is what makes lithium-ion flammable. It is also one of the things limiting how much energy you can safely pack in because if we go above the limit fire breaks out.

Solid-state batteries replace that liquid with a solid material which could be ceramic, glass, or some other polymer. This one change has many differences for the battery.

No liquid means no flammable material. No thermal runaway. No phone fires. No EV fires that burn for hours because the liquid electrolyte keeps feeding the flame.

Solid electrolytes also allow for a lithium-metal anode instead of graphite. Lithium metal can store ten times more energy than graphite.

Combine that with the solid electrolyte and you get energy densities that graphite-based lithium-ion cannot come close to.

Just for reference, today’s best lithium-ion cells sit around 300 Wh/kg.

Toyota’s solid-state batteries in development are hitting 450 to 500 Wh/kg. BYD’s solid-state cells have achieved 400 Wh/kg in testing. CATL is targeting up to 500 Wh/kg.

Toyota is also claiming that charging for ten minutes would give a range of 1,200 kilometers in an EV. That is not a typo.

Big manufacturers like BYD, Toyota, CATL, are all targeting small-scale production around 2027 with broader rollouts closer to 2030.

Toyota’s solid-state batteries even received production approval in Japan in late 2025.

The solid-state battery technology is already here and it’s only a matter of a couple of years when we start seeing them used starting with electricity-hungry EVs and then to our peasant devices :p

Why Is Solid-state Not Everywhere?

The obvious question is if solid-state is so much better, why is it not in your phone right now?

Because building solid-state batteries right now is extremely hard, expensive and phone customers aren’t asking for it right now.

For a more technical answer:

The solid electrolyte and the electrodes inside a solid-state battery need to stay in perfect physical contact with each other through hundreds of charge and discharge cycles.

In a liquid system, this happens naturally - liquid fills every gap. In a solid system, tiny imperfections in contact mean ions cannot move properly, and the battery stops working.

Manufacturing solid-state cells requires completely different production equipment from everything the industry has already built for lithium-ion.

Factories that make lithium-ion batteries cannot simply retool. New facilities, new machinery, new processes - all of that costs enormous amounts of money and time.

And then there is the cost.

The material cost of solid-state batteries right now is estimated at $400 to $500 per kWh, compared to around $100 per kWh for standard lithium-ion. That gap has to close before solid-state can show up in mainstream phones and affordable EVs.

A senior scientist of battery company CATL said that the technology’s maturity is currently a 4 out of 9. They want to get to 7 or 8 by 2027.

To put it simply:
For solid-state batteries the whole supply chain needs to be built around solid-state tech and building that would be slow and expensive, although it’s inevitable because the tech is real.

The Next Decade Of Batteries

For me personally, the next decade is going to be really interesting from the technology POV.

Lithium-ion batteries have been with us for quite some time ever since Sony first commercialized them back in the 90s.

In that time, lithium-ion batteries have not just drastically improved but they’ve also evolved into better versions like Lithium Iron Phosphate (LiFePO4) which is the latest standard that is getting more dominant.

The purpose of this piece was to discuss the emerging technology that is being built in front of us and while we’ve talked about the tech and some estimates from experts please take them with a grain of salt.

The industry believes that Silicon-carbon batteries are going mainstream between now and 2030. Your next phone will have a noticeably better battery than your current one without the device getting any bigger.

Charging speeds will also get faster because silicon anodes handle higher charge rates better than graphite.

Solid-state batteries will first appear first in niche applications like high-end EVs, possibly some wearables and drones before gradually moving into mainstream consumer electronics.

Another angle that makes me happy about improving batteries is a local one.

See, Pakistan has an electricity grid problem. I won’t go into much detail but the fact is that our grid-electricity just sucks. There are constant shutdowns, load-shedding and whatnot.

UPS has been a solution forever but with China’s incredible success in producing lithium-ion batteries as well as ever-efficient solar panels, Pakistan has been among the biggest importers.

Houses that can afford the system have given up the government’s unreliable grid and have flooded their rooftops with as many solar panels as they could afford.

Just take a walk in any big city and you’d notice quite a lot of panels. With that the affluent ones have installed batteries in capacities that can power their entire households’s electricity requirement.

That’s thanks to the competition and innovation that China brought to lithium-ion that we’re able to enjoy batteries that now last more than a decade on average costing a couple hundred thousand rupees.

And the rising competition between the US and China for the next generation of battery technology is only to be seen as the years go by.

The Pakistani government has also taken a note. Just a few months ago they released their five-year policy that focuses on lithium-ion cells (foundation of lithium batteries) being produced locally which atm are totally imported from China.

All the major companies in the country are launching their versions of lithium batteries. Right now, the lithium cells are totally imported but with time we can see that localize here in Pakistan.

Maybe I can write on that when we have some good progress on this front. If that’s something that excites you, do let me know down in the comments.

More from SK NEXUS

As always I’d love your thoughts on my writing or your questions around battery technology. If you have any feedback to share please feel free to engage in the comments down below or make your voice heard in a fun poll down below:

Further Learning

• The Age of Silicon Is Here…for Batteries

• The EV Battery Tech That’s Worth the Hype, According to Experts

• The age of HUGE batteries is here!

• The Problem with these Smartphone Batteries

• The Evolution of Batteries: Voltaic pile to Lead-acid

• Silicon Carbon Battery in Smartphones: Explained

• Batteries Still Suck, But Researchers Are Working on It

• CES 2026: The First Solid-State Vehicle May Be a Motorcycle

• Pakistan’s surprise solar surge shocks experts and grid

Looks like you made it to Part 4, I respect the consistency.

Quick recap - in the last piece, we focused on browsers aka the primary gateway to surveillance. I broke down how third-party tracking works, why default configurations aren’t enough, and how to harden your setup properly. If you skipped that one, go back and read it first. This builds directly on it.

For anyone new here - this is Part 4 of a practical privacy series designed to help you secure your digital life step by step.

Now, up until this point if you have done everything right. That still doesn’t mean you are truly protected because your IP address is exposed and tied directly to you, that’s still a wide attack surface.

Your IP address is your rough digital location. It’s assigned by your Internet Service Provider. And even with HTTPS, your ISP can see a lot because metadata still exists, DNS requests still travel and connection records still get logged.

So the question becomes simple:

If the network itself sees you, how do you reduce that visibility?

That’s exactly what we’re breaking down in this part.

And just a heads up for the new folks - if you’re unsure about the order of this series, follow it like this:

• Privacy Is Not Security - The Difference is Important

• Privacy Fundamentals - Part 1: Authentication

• Privacy Fundamentals - Part 2: Communication

• Privacy Fundamentals - Part 3: Your Browser

The Core Issue: Network-Level Exposure

When you go online. You are basically moving through infrastructure and whoever controls that infrastructure, is the big guy.

Start with your Internet Service Provider.

Every request you make - passes through their network. Even when traffic is encrypted via HTTPS, your ISP can still see the domains you connect to, your destination IP address, session duration. This is called metadata about which we previously discussed as well.

Similarly, let’s say you are sitting in a coffee shop or maybe an airport and using the public Wi-Fi.

By doing that - you’re plugging your device into a network you don’t control, administered by people you don’t know, and shared with strangers. HTTPS protects the content of most modern websites. That’s good.

But it doesn’t make the environment trustworthy.

The person operating the network can still see where you’re connecting. They can attempt traffic manipulation. They can redirect users to phishing pages and if any connection on your device isn’t using an encrypted protocol, then they can take advantage of it.

The main problem is that you are operating inside infrastructure that isn’t yours.

The same applies for different social media companies, content platforms, cloud providers - your data flows through their servers. They sit at the main pathways and whoever controls those pathways controls information flow.

That control isn’t automatically malicious. But structurally, it grants power to log, to block and to do much more.

Understanding VPNs

A VPN aka Virtual Private Network is one of the most marketed privacy tools on the internet.

At its core, it creates an encrypted tunnel between your device and a remote server operated by the VPN provider. Instead of your traffic going directly from your device to the website you’re visiting, it first travels through that encrypted tunnel to the VPN server and only then out to the wider internet.

In practical terms, this adds a protective layer. But it’s important to understand what that actually means, and what it doesn’t.

How a VPN Works

There are three core functions happening when you connect to a VPN:

• Encryption - Your internet traffic is encrypted before it leaves your device. Anyone monitoring the local network, including your ISP or a public Wi-Fi operator - sees scrambled data rather than readable content.

• Secure Data Transmission - On compromised or public Wi-Fi networks, this encryption becomes critical. Even if someone intercepts your traffic, they can’t easily break it.

• IP Address Masking - When connected, your visible IP address changes to the IP of the VPN server. To external websites, it appears as if your traffic is originating from that server’s location instead of your real one.

That masking is what makes it appear as though you’re browsing from another country.

What a VPN Actually Hides

IP Address and Geolocation

For those of you who don’t know, an IP address (Internet Protocol address) is a unique numerical label assigned to your device whenever it connects to a network. It identifies your device and provides a routing location for internet traffic.

Without a VPN, that IP address is visible to websites, trackers, and your ISP.

With a VPN, the IP assigned by your ISP is hidden from the websites you visit. Instead, they see the VPN server’s IP. That means:

• Your real location is masked

• Trackers can’t directly tie browsing sessions to your residential IP.

• Your ISP sees encrypted traffic going to the VPN, not the final websites you access.

If you’re curious what your current IP reveals, you can visit this link to do that.

And yes - this is also why VPNs are commonly used for bypassing regional restrictions. Let’s say if you’re in the USA and a Netflix title isn’t available locally, connecting to a server in another country can make it appear as though you’re browsing from there.

Search Activity and Torrents

When using a VPN properly - your ISP cannot see the specific websites you visit. Your local network observers cannot read your traffic and your browsing activity becomes harder to link directly to your home IP.

If you avoid logging into identifiable accounts unnecessarily, correlation becomes more difficult. Trackers still exist - but without your direct IP address, cross-site tracking weakens. Your browsing habits are less easily tied back to your physical connection.

Overall, a VPN does a solid job of protecting a significant amount of personally identifiable network data.

But that’s not the whole story.

What a VPN Does Not Hide

This is where people get misled. It’s important to note that a VPN does NOT hide your digital fingerprint, your logged-in personal accounts, any kind of malware infections or protect you from getting social engineered.

If you sign into Google or any personal account, those platforms know exactly who you are, it doesn’t matter if you are using VPN or not, at that moment. Or if your browser fingerprint is unique, websites can still track you or if you download malware, a VPN won’t stop it.

A VPN is not a bulletproof vest, it’s a layer and like every layer we’ve discussed in this series, it only works properly when combined with disciplined behavior.

Tools matter, but habits matter more.

VPN vs WARP vs Proxies

VPNs, WARP, and proxies are often used interchangeably online. They shouldn’t be. They solve different problems, operate at different layers, and carry different trade-offs.

If you don’t understand the distinction, you’ll either overestimate your privacy or use the wrong tool entirely.

WARP

Cloudflare WARP encrypts your DNS requests and routes your traffic through Cloudflare’s network. The goal is to protect you from local network snooping and improve connection reliability using Cloudflare’s infrastructure.

It does a good job securing traffic on public Wi-Fi and preventing your ISP from easily seeing specific domains you connect to. But it is not built primarily as an anonymity service and it does not prevent Cloudflare from having visibility into traffic flowing through its own network.

In short, WARP strengthens connection security and is not a serious anonymity layer.

Proxies

A proxy is the simplest of the three.

Instead of your device connecting directly to a website, the connection goes through a middle server first. The website sees the proxy’s IP address, not yours.

That’s the core function.

Unlike a VPN, a proxy typically does not encrypt your traffic end-to-end. That means your ISP can still see what you’re accessing unless additional encryption (like HTTPS) is in place. Proxies are commonly used for automation, bypassing regional restrictions, or routing specific applications through alternate servers.

In short, Proxies are functional tools, not privacy guarantees.

These technologies may look similar from the outside. Under the hood, they are solving different problems. Here’s a table to make the difference more prominent:

The VPN Trade-Off

You thought there’s no trade-off? Here’s an eye opener:

When you activate a VPN, your ISP mostly sees an encrypted tunnel going to one server. They no longer easily see the final destinations of your traffic.

But now the VPN provider becomes the one in control.

Technically, that VPN provider can see your real IP address when you connect. They can see which VPN server you’re assigned. They can see connection timestamps. Whether they log that information depends on internal policies.

And this is exactly why logging policies, jurisdiction and ownership matters.

A subpoena aka a legal order requiring a company to provide records - can compel a VPN provider to hand over whatever data they have retained. There’s a real-world example of this:

In 2017, U.S. authorities were investigating a cyberstalking case. They traced suspicious activity to an IP address belonging to PureVPN. The company had advertised a “no-logs” policy, but it turned out that wasn’t the case. When served with a subpoena, PureVPN provided the data it had. That information helped identify and arrest the suspect.

The deeper issue wasn’t that they complied with a lawful request - most companies will. The issue was transparency versus marketing language.

One last thing, don’t even think about using free-VPN’s. Servers, bandwidth, and infrastructure cost money. If a VPN is free, revenue comes from somewhere. In many cases, that somewhere is user data getting sold.

Keep in mind - if something is free, you are often the product.

What Should You Actually Use?

So, whether you’re on public Wi-Fi at a café, w or just scrolling on your phone, the same basic principle applies - your internet connection reveals a lot about you, your IP address alone can expose your approximate location, the network you’re using, and create a persistent identifier tied to your activity.

Now, the question is for a casual internet user who simply wants better privacy and safer browsing habits, what can he/she do? Well, here are three realistic options:

Mullvad VPN - The Most Transparent Option

If your goal is serious privacy at the network level, Mullvad is one of the strongest options currently available.

Unlike most VPN providers, Mullvad does not require an email address or personal details when creating an account. Instead, it generates a random account number that becomes your only identifier.

Jurisdiction also plays a role here. Mullvad operates out of Sweden, where privacy protections and regulatory structures are relatively strong. While many companies claim a no-log policy, Mullvad has repeatedly positioned itself around minimal data collection by design - reducing the amount of information that could exist in the first place.

Another feature that privacy-conscious users appreciate is payment flexibility. Mullvad accepts several forms of payment, including cryptocurrencies, which means users can avoid attaching card details if they prefer.

Although the service is not free, the pricing is intentionally simple - one flat monthly rate regardless of how long you subscribe.

If you want to explore the deeper technical breakdown, this review provides additional details: https://cyberinsider.com/vpn/reviews/mullvad-vpn/ And if you’d like to get started directly, Mullvad provides setup guides for every platform here: https://mullvad.net/en/download/vpn/linux

Proton VPN - A Reasonable Free Alternative

For people who are not ready to pay for a VPN service, Proton VPN is one of the few free options that still maintains a strong privacy reputation.

Proton is widely known for building privacy-focused software products, including encrypted email and secure cloud services. Their VPN follows the same philosophy.

Even the free version of Proton VPN maintains a fairly strong privacy posture. However, there are limitations.

But many advanced features are reserved for paid subscribers. This is typical across most VPN services. Another downside common to free VPN tiers is traffic overload on servers. Because thousands of users rely on the same limited infrastructure, free servers can become overloaded, which impacts connection speed.

Despite these limitations, Proton remains one of the few providers offering a free service that does not appear to rely on aggressive data monetization.

And yes, proton also has a paid version, here are the details for it:

For anyone exploring VPNs without committing to a paid subscription immediately, Proton VPN is one of the safer starting points - you can download it by visiting this link.

Cloudflare WARP - A Lightweight Privacy Layer

If you are someone who believes they don’t need the additional features that a full VPN provides. And If your primary goal is simply preventing an ISP or local network from easily monitoring traffic - Cloudflare WARP can be a practical alternative.

WARP works by routing your internet traffic through Cloudflare’s network while encrypting DNS requests and connection data. Instead of exposing your original IP address to websites, your traffic appears to originate from Cloudflare’s infrastructure.

However, it’s important to understand what this does and what it does not do.

WARP replaces your public IP address with a Cloudflare IP that still reflects your approximate geographic location. It is not designed as an anonymity service and does not attempt to conceal your location in the same way traditional VPN services do.

For users who want something lightweight that improves connection security without managing a full VPN service, it can be a convenient solution.

You can download the client directly from Cloudflare here: https://developers.cloudflare.com/warp-client/get-started/windows/

The Habit Side of Privacy

If you have made it this far, it means you have addressed several major layers of digital privacy including identity and authentication, secure communication tools, browser choices and tracking behavior. And now we’ve looked at the network layer - where your IP address and connection metadata exist.

Each one of these layers reduces exposure in a different way.

Understand that using a VPN does not suddenly make someone anonymous online and that privacy is not a single tool you enable once and forget. It’s a set of habits built gradually over time. The tools help, but behavior ultimately determines how effective those tools are.

That’s why stacking everything you have learnt is where meaningful protection emerges.

By the way, there is still one layer we haven’t explored yet - the operating system itself.

If privacy is your priority, know that Microsoft Windows is not something worth using and the natural alternative is Linux.

If enough of you are interested, the next article can break that down in detail - including why Linux is often the better choice for privacy, which distributions make sense for beginners, and how to transition without breaking your workflow.

For now, this wraps up the series.

This was the final part, but it builds on everything that came before it. The previous four parts break down the foundations in detail, so if you haven’t gone through them yet, it’s worth starting there.

Thanks for reading.
See you in the next one.

One Last Thing

Nearly all people interact with systems without questioning them. This space is about doing the opposite aka understanding how things work, where they fail, and what that means in practice.

If that’s the kind of content you want more of, subscribe. And if this piece added something new to how you see things, pass it on. Good information only matters if it travels.

And if you’re active elsewhere, follow along on Bluesky or X (Twitter) or LinkedIn - those are the two places I show up most consistently outside Substack. A follow, a like, a repost - it matters to me, more than you think.

What You Might Like

Let me tell you something that happened to me recently.

I was sitting (as usual) at my 2019 Dell workstation, writing, doing some research, having a bunch of tabs open. The usual stuff. Everything was running fine.

And then a notification popped up that this new shiny processor just got launched by Intel (It was the Core Ultra 270K btw).

I sat there for a second but then I realized that my work is going fine with my laptop and computer that I already have. Even if I had the money, would it be reasonable to buy the latest processor just because it’s cool tech?

In 2026, keeping your old hardware is not just acceptable. In many cases, it is the smarter move. And I think a lot of people need to hear right now.

We are seeing companies like NVIDIA bringing back their old cards like 3060 back into production because they know too that AI crises has raised prices by alot.

On the consumer side, we’re also seeing people buy used hardware or hardware that is a few years old. The most sold CPUs are AMD processors which are 2-4 years old.

And all the while Linux is making headlines into consumer devices who are sick and tired of Microsoft Windows. Because just a few days ago, Steam’s March 2026 stats show Linux doubling it’s market share from a few months ago, in December 2025.

As things stand, Linux has outpaced MacOS numbers. From the March 2026 report, Linux now stands at 5.3% while MacOS stands at 2.3%.

And actually, there’s nothing to be embarrassed about if you’re not upgrading to the bleeding-edge tech right now. It’s the smarter move especially we have wars all over our world and AI too.

In this article I’ll share my experience buying used tech hardware and discuss why in many cases it makes sense to buy used tech hardware esp in country like ours where we earn in PKR and spend in USD.

The Upgrade Itch

You know the feeling. At least I know it ever since forever.

A new chip drops. A benchmark video shows up in your feed. Someone in your circle gets a new laptop. And suddenly your perfectly working machine starts to feel slow. Outdated. Behind.

It is not real. But it feels real. And it is supposed to feel that way.

Tech marketing is specifically designed to create that itch. Every new chip is “the fastest ever.” Every new laptop is “twice as fast as the previous generation.” And technically, a lot of that is true.

But here is what the benchmarks never tell you. Twice as fast at what?

If you are not doing the thing being benchmarked, the speed difference is invisible in your daily life. Opening Chrome does not feel faster. Writing does not feel faster. Watching a video does not feel faster.

The gap between what your machine can do and what you actually need it to do is wider than it has ever been for most people.

That gap is the whole argument.

AI Made Everything Expensive

Here is the plot twist nobody saw coming a couple of years ago.

AI did not just change software. It changed hardware prices in a way that hits every regular person trying to buy or upgrade a computer right now.

DDR5 RAM prices jumped 60 to 75% in just a few months through late 2025. 16GB modules that were selling for around $7 to $8 earlier in the year climbed to roughly $13 by the end of it. And that is the mild version of the story.

Some specific kits saw increases of 500%. A 32GB DDR5 kit that cost around $87 in early 2025 was listed at $399 on Amazon by mid-December.

Why? Because the companies building AI infrastructure - your Microsofts, your Metas, your Googles - need enormous amounts of memory for their data centers. And the three companies that make almost all of the world’s RAM - Samsung, SK Hynix, and Micron - started redirecting production toward high-bandwidth AI memory because it is far more profitable.

Memory manufacturers have reallocated more than three times their wafer capacity toward HBM chips for AI workloads. That leaves fewer DDR4 and DDR5 chips for regular consumers. Basic supply and demand.

NVMe SSDs followed the same trajectory. NAND flash prices went on an almost identical ride.

Industry analysts are warning we have not hit peak pricing yet, with meaningful relief possibly not arriving until late 2026 at the earliest.

So this is genuinely not a great moment to be buying new hardware.

And conveniently, as we are about to discuss, it is also not a necessary one.

If you want to learn more about how AI made our tech expensive, here’s a detailed article I did here on SK NEXUS.

A Reasonable Upgrade Cycle

Here is something the tech industry does not want you to think about too hard.

PC hardware has historically moved in roughly three-year meaningful performance cycles. Every few years there would be a genuine leap that made a real difference to everyday users.

But lately those genuine leaps have been narrowing for most workloads. The jump from one chip generation to the next delivers impressive benchmark numbers.

The actual experience for someone doing documents, browsing, video calls, and light creative work is increasingly difficult to distinguish from the previous generation.

Think about what actually triggers a genuine need to upgrade.

A piece of software you rely on stops running on your machine. Your hardware physically fails. You start doing a genuinely new kind of work that your current machine simply cannot handle.

For most people, none of those triggers have fired in a long time.

The software still runs. The hardware still works. The workload has not changed dramatically.

The upgrade pressure you feel is real. The reason behind it mostly is not.

The MacBook M1 example

Apple released the M1 chip in late 2020. It was a genuine generational leap - fast, efficient, ran everything well.

Here we are in 2026. The M4 and M5 are both out now. Benchmarks are impressive. The M5 is roughly 2.5 times as powerful as the M1 on certain workloads.

And yet a huge number of people are still on M1 machines. And they are completely fine.

Writers are writing on them. Designers are designing on them. Developers are coding on them. The machines have not become slow. The software has not stopped working.

The experience has not degraded.I personally know people who are fine with their M1 MacBook Pros.

This is what a good hardware purchase looks like across time. You buy something solid, you use it, it keeps working, and you do not upgrade until there is a real reason to.

The M1 owners still running their machines in 2026 did not make a mistake. They made a smart purchase and are getting full value out of it years later.

I am living a version of this myself. My 2019 Dell workstation handles everything I throw at it without complaint. Writing, research, video calls, light editing, multiple browser tabs. All of it. No issues.

The machine is not the bottleneck in my life. My workflow is.

Faster hardware would not make me faster. I had to be honest with myself about that.

The Windows 11 Problem

I don’t want to start my rant on Windows 11. Just yesterday I found 15 minutes trying to find an option to troubleshoot my bluetooth headphone.

I wanted to play a game and the audio sucked on Bluetooth because of a setting that I had to change. Turns out, I had to go through 3 different settings pages just to get to that.

That’s Windows 11 for you. But this isn’t the only big problem. A major problem was Windows 11’s launch.

Back when it launched, Windows 11 came with a set of hardware requirements that locked out a massive number of perfectly working machines.

The two big ones are TPM 2.0 - a security chip requirement and a specific list of supported processors up to 2016-era.

Computers from before, machines that run everything fine, were officially declared incompatible.

Owners of older but capable systems face a barrier. Their hardware may run Windows 11 perfectly except for lacking a TPM or being outside the CPU compatibility list.

That is not a a hardware problem even though it’s presented as a security problem.

Microsoft’s restrictions are estimated to threaten the usability of hundreds of millions of existing devices.

Windows 10 support officially ended last year and a massive number of functioning computers were quietly pushed toward replacement.

Environmental groups called this what it is. Planned obsolescence. The machine works. The software says it does not. Those are two very different things.

My fellow wrote an in-depth exploration of this concept of Planned Obsolescence in his article that you could checkout here.

But here is the thing. This whole situation has a solution. And that solution has been getting better for years.

Linux Becoming Accessible

Before you get into this section, if you don’t know what Linux is or need a simple refresher, you may go through my article linked here where I explain what Linux is and why you should care.

Full disclaimer. Linux used to be genuinely difficult. And that’s coming from someone who’s been daily driving Linux since 8 years now.

It was a hobbyist operating system that required patience, comfort with the terminal, and a high tolerance for things occasionally not working. For most people it was not worth the trouble.

That has changed. Significantly.

In 2021, Linus from Linus Tech Tips - one of the biggest tech YouTube channels in the world did a series where he tried to use Linux as his daily driver. It was called “Linux Hates Me” which gives you a sense of how it went.

Linus ran into real friction. Things did not work the way he expected. He got frustrated. His co-host Luke had a much smoother time on Linux Mint. The series ended up being an honest portrait of where Linux stood then. Capable, but not quite plug-and-play for everyone.

That was 2021. It is 2026 now. And you know what? Linus is doing another one-month Linux challenge. That shows the renewed interest in Linux.

Modern distributions like Ubuntu, Fedora, and Linux Mint have closed most of the gaps that made everyday use difficult. Installation is straightforward. Most hardware is recognized automatically. Browsers work. Office suites work. Video calls work.

Gaming on Linux through Valve’s Proton compatibility layer has gotten to the point where a large percentage of Steam games just run, without any extra work from the user.

For someone whose machine has been locked out of Windows 11, Linux Mint is probably the most sensible next step. It looks familiar. It runs on older hardware beautifully. It handles the daily tasks that most people use a computer for. You do not need to know terminal commands. You just install it and use it.

It is not perfect. Some software does not have a Linux version. Some workflows need Windows-specific tools.

But for a large percentage of people whose machines Microsoft has declared obsolete, Linux is not a compromise. It is an upgrade.

PKR Earnings, USD Spendings

Let me get a little local for a second because you won’t hear this discussion in an article for US audience.

Everything we have discussed so far - the AI-driven RAM price surge, the cost of upgrading, the reasons to hold on to old hardware, all of that hits differently when you live in Pakistan. Or anywhere in South Asia, honestly.

Here is the basic problem. Almost every piece of tech you buy in Pakistan is imported. Laptops, GPUs, SSDs, phones - nearly all of it comes from China, Taiwan, the US, or the UAE. And all of it is paid for in dollars.

The dollar rate in 2025 was hovering around PKR 280, fluctuating with political instability and global economic pressure. Even a small move in the exchange rate directly changes what you pay at the counter. The machine did not get more expensive overnight.

The rupee got weaker. Same thing, different cause. And then there are the taxes on top of that.

Mobile phones imported into Pakistan are subject to taxes too, an 18% sales tax, withholding tax, a handset levy, and regulatory duties, all stacked on top of each other.

The FBR doubled the sales tax on imported computers and laptops in the 2024-25 tax year. The customs valuation system that determines how much duty you pay has been revised multiple times, sometimes helping, sometimes making things worse.

The result is that a laptop that costs $800 in the US does not cost the equivalent of PKR 240,000 here.

By the time import duties, sales tax, distributor margins, and retailer markup are applied, you are sometimes paying 30 to 50% more than what someone in the US or Europe pays for the exact same machine.

And unlike India , which has started building local assembly plants for brands like Dell and Acer to bring costs down - Pakistan has no major domestic tech manufacturing to speak of.

Although things are changing here. There are some companies doing some level of assembly here for TVs and smartphones but that’s still nowhere near as big as India.

Generally, every single laptop on the shelf at your local shop went through the full import chain. Every price you see carries that overhead.

This is why the “just buy new hardware” advice that floats around Western tech YouTube makes no sense here.

Linus can upgrade his workstation every two years and it is a reasonable decision for someone in Canada with Canadian prices and Canadian salaries. For a student in Lahore, a decent mid-range laptop can easily cost two to three months of salary.

You do not treat that purchase casually. You do not upgrade on a whim. You use what you have for as long as it works, and you squeeze every bit of life out of it.

That is not being cheap. That is being rational.

And it is exactly why old hardware makes more sense here than almost anywhere else.

Your 2019 machine is not just running fine. It is running fine and it would cost you an arm and a leg to replace it right now with something meaningfully better.

Hold on to it.

When Upgrading Actually Makes Sense

Okay. I have spent most of this article making the case for old hardware. Let me be fair and give you the honest answer to when upgrading is genuinely the right call.

There are real triggers.

A piece of software you depend on for work stops supporting your hardware entirely, with no workaround. Your hardware physically fails and repair costs more than replacement.

Or you take on a genuinely new workload - running local AI models, editing 4K video on a machine without the processing headroom, serious 3D rendering work.

Those are real reasons. Everything else is probably the upgrade itch talking.

And if you are not in one of those situations but your machine feels slow, there are cheap moves worth trying first.

Adding RAM if you are on 8GB is still the single most impactful cheap upgrade for most people. An SSD if you are still on a spinning hard drive transforms how a machine feels - used SSDs are still affordable right now.

A clean OS install often fixes half the sluggishness people blame on aging hardware.

And if your machine has been locked out of Windows 11, try Linux Mint before you decide you need to buy anything.

You might be surprised.

How I’m Using Used Tech

Let me just tell you where we are.

RAM and SSD prices are up there and probably will not correct significantly anytime soon.

New hardware is more expensive than it was a few years ago and the performance gains for everyday workloads are incremental rather than disruptive.

I covered this concept of progressive and disruptive in a detailed article which you can read here.

Your existing computer, if it was decent when you bought it, is probably still decent now. The software still runs on it. The work still gets done on it.

Look, I personally know people who bought a latest-gen machine back in 2019 or so and are still rocking that machine because their workflow hasn’t demanded more.

Heck, Even I’m using a computer from 2019 that is perfectly fine for my workload as it stands currently. I’m literally using that same computer I’m writing this article on :p

If you want to know how my 2019-era computer is working great, you can checkout the full article I wrote on it last year.

So, The upgrade pressure you feel is real but it comes from marketing cycles and product announcements, not from your actual experience of using your computer.

Your decision should not be entirely based on that. Because if you do there is no stop to it. You should be able to take a deep breath and analyze thing for yourself.

Basically, don’t just give into the hype and do some research when it comes to your usecase and expectations from the tech you buy.

Hold on to what you have. Wait for prices to correct. Only buy something new when you genuinely need it, not when Apple launches a new MacBook.

Your 2019 machine might have more life in it than you think.

Mine definitely does. And I’m happy with it even though everyday my body is itchy when I discuss new technology that is coming out.

Further Learning

• Everyone is Buying Old Hardware

• NVIDIA RTX 3060 Production is Restarting...

• These real-world tests will convince you to switch to the M5 MacBook Pro

• Do You Need to Upgrade Your MacBook Air?

• Microsoft’s draconian Windows 11 restrictions

• How Did RAM Get So Expensive? And How it’s Going to Get Worse…

• Linux Just Hit a Big Milestone in the Desktop OS Race

• Why people keep flocking to Linux in 2025

If you’ve consumed a minute of tech news the last few weeks you have surely heard of that $599 MacBook that has all the tech world by storm.

And by now you might be somewhat bored by seeing the MacBook Neo over your Youtube, Instagram or Tiktok feed even.

Look, it’s common for Apple products to receive special spotlight at launch but the recently launched MacBook Neo deserves the high praise.

And it’s not just me saying it. Linus Sebastian from Linus Tech Tips, being someone who is somewhat critical of Apple products at times, has said that he can only recommend MacBooks now.

If you had told anyone in the tech community that Linus would say that in 2018, they would have laughed at you.

Back then the arguments for buying Windows were genuinely stronger. MacBooks were expensive, ran hot, had bad keyboards, and the performance for the money just wasn’t there.

So what happened between then and now? And why is a $599 laptop the clearest and most surprising expression of it yet?

So, this isn’t another MacBook Neo article but it’s one that covers what was wrong with Apple before their own Silicon.

I’ll talk about what Apple did that no-one else could that led to we getting a $599 laptop from Apple.

And of course there will be a comparison of Neo to laptops at its price-point.

MacBooks Before Apple Silicon

In 2026, I feel if you have to talk about MacBooks you can divide all of them into two categories; Before Apple Silicon and After Apple Silicon because that’s what has provided the biggest edge to MacBooks.

Today’s MacBooks look perfect but it’s important to know that they weren’t like this forever.

To understand why the MacBooks of today are so good, let’s discuss some of the issues that plagued MacBooks before Apple Silicon arrived on the scene.

The Butterfly Keyboard

So, let’s start with the butterfly keyboard.

From 2015 to 2019, Apple shipped MacBooks with something called the butterfly keyboard. The name sounds beautiful but the product really was not.

The butterfly mechanism was an ultra-thin keyboard design Apple used across its entire MacBook lineup for four years running.

The pitch was that thinner keys meant a thinner laptop. The reality was that if a single crumb, a grain of dust, or a tiny piece of debris got under a key, the keyboard would stop working.

Keys would repeat on their own without being pressed. Letters would randomly stop registering. Some people had multiple keys fail simultaneously.

Apple’s solution was to replace the broken butterfly keyboard with another butterfly keyboard. The exact same design. And it would break again.

Reviewers questioned its reliability publicly. The keyboard became a running joke in tech circles. And then it stopped being funny when people started filing lawsuits.

Apple eventually paid $50 million to settle a class-action over it. But the keyboard was just the beginning.

The Dongle Era

The 2016 MacBook Pro redesign removed every port except USB-C. Not some ports. Every single port.

No USB-A. No HDMI. No SD card slot. Just USB-C holes and nothing else.

If you owned an external hard drive, a USB drive, a camera, or a monitor - all of which most people who spend $1,300 on a laptop tend to own - you now needed a dongle to use them.

The dongle era became its own meme. People sitting in coffee shops with a cluster of white adapters hanging off their MacBook like a bunch of grapes.

Apple is selling a $9 dongle for a machine that cost $1,299. Reviewers called it the most dongle-dependent laptop Apple had ever shipped.

The Heat Issues

And then there were the heat issues.

MacBook Pros from 2018 and 2019 showed severe throttling issues. The 2018 15-inch model launched with a software issue that slowed the processor below its rated speed under load, and Apple had to send out a formal apology.

Even though Apple sent patches the problem did not fully go away because the root cause was not the firmware, it was the build itself.

Apple also had to bear the brunt of Intel’s failure because back in 2015 Intel promised delivery of 10nm chips well before they actually arrived.

Apple had already designed the 2016 MacBook Pro enclosure around those promised chips, and Intel was years late.

The heat piping, fan design, and laptop dimensions were all built around a processor that never showed up on schedule, leaving Apple cramming more powerful Intel processors into a body that was too thin to cool them properly.

What this meant in practice: the CPU would hit 100 degrees Celsius under load and throttle down to protect itself.

A $2,000 MacBook Pro with an Intel i9 processor - Apple’s most expensive consumer laptop - was slower under heavy workloads than the i7 model because the i9 generated more heat than the chassis could handle.

Dave Lee, a popular tech reviewer at the time, demonstrated this on YouTube and it went viral. A $2,300 laptop throttling itself below its advertised speeds was not a minor software bug. It was a fundamental design problem.

The 2017 MacBook Pro still carried the defective butterfly keyboard and had performance troubles from thermal throttling. CPUs slowed under heat, undermining the “Pro” status, and the faster Kaby Lake chips barely mattered in practice.

The Pricing Issues

And the price. Let’s talk about the price.

The base 2017 MacBook Pro with Touch Bar started at $1,299 and came with a dual-core Intel Core i5, 8GB of RAM, and 256GB of storage.

In 2017, you could buy a Windows laptop with a quad-core processor, 16GB of RAM, a dedicated GPU, a full set of ports, and a backlit keyboard for $700.

The MacBook cost nearly double, offered half the RAM, had a keyboard that might stop working if you ate a biscuit near it, ran hot, and required a separate $50 dongle.

High starting price relative to Windows alternatives with similar raw specs, accessories and dongles increasing total cost.

Many users found the premium unjustified given compromises in ports, keyboard reliability, and thermal limits.

The arguments for buying Windows during this period were not just valid. They were overwhelming.

Although not everything was bad with the MacBooks. There were things the Intel MacBooks did well.

For example, the display quality was excellent - Retina panels with P3 wide color gamut that Windows laptops at the same price could not match.

Build quality was still best in class. macOS was (and still is) a genuinely better operating system for a lot of people’s daily workflows.

The trackpad was the best in the industry by a wide margin. And for light work - browsing, writing, email, the machines were perfectly capable.

But for anyone spending $1,500 to $2,000 on a machine they expected to perform like a professional computer, the Intel era MacBook Pro was a consistently disappointing experience.

The gap between what it cost and what it delivered was hard to square. And Apple knew it.

The thermal issues in the MacBook line were likely one of the main reasons Apple moved strongly toward its own cool-running ARM processors.

Apple’s move to ARM has been great for the company. If you’re not up-to-date on what ARM means in this context, you may checkout my article linked here.

The decision to ditch Intel was not just about ambition. It was about fixing a problem Apple could not fix any other way.

The chassis was too thin for Intel’s chips of the time. And Intel kept missing its own deadlines for better, more efficient chips.

Apple had been waiting years for Intel to deliver something it could actually cool properly, and Intel kept disappointing them.

So Apple built its own.

How Apple Silicon Broke The Laptop Industry

In 2020, Apple announced it was done with Intel.

After fifteen years of using Intel chips in every single Mac, Apple decided it was going to build its own processor. Their first chip was called M1. And it landed like a bomb.

The MacBook Air with M1 launched at $999 with battery life that lasted well over twelve hours in real-world reviewer testing and not just in Apple’s controlled lab conditions.

The machine ran completely silently because it had no fan inside at all. It was faster than most Windows laptops that cost twice as much.

And the surface of the machine barely got warm even under load compared to every Intel MacBook that came before it.

People who had been dismissing Macs for years suddenly had very little to say.

Here is the key thing to understand about why M1 was so fundamentally different from what came before it.

Traditional laptop chips - your Intel and AMD processors are built from separate components that communicate with each other over something called a bus.

The CPU is one piece. The GPU is another piece. The RAM is separate from both.

When any of these components need to share data (which they do constantly) the data has to travel across that bus. That travel creates latency using energy while generating heat as a result.

Apple’s M1 put everything on a single piece of silicon. The CPU, the GPU, the memory, the neural engine for AI tasks - all of it sits on one chip and shares data through a unified RAM.

There is no bus for data to travel across between these components because they are all right next to each other.

The result is a system that does significantly more work with significantly less energy, generates much less heat, and on everyday tasks runs faster than Intel chips that consume two or three times as much power.

Reviews of the M1 MacBook Air showed that in terms of thermals under normal workloads, the machine remained comfortably cool while performing better than Intel-based Macs.

The passive cooling (no-fan design) turned out to barely matter for the tasks most people actually do on a laptop.

Browsing, watching videos, video calls. The M1 handled all of it without breaking a sweat and without making a single sound (because no fan, remember?).

Battery life on the Intel MacBook Air before M1 was around 6 to 8 hours depending on what you were doing.

The M1 Air took 12 to 15 hours. That is not a marginal improvement. That is a completely different category of machine.

And Apple didn’t stop there but kept going.

M2. M3. M4. M5. Each generation built on the architecture of the last. Each generation compounded the advantage over Windows.

Each year the gap between what a Mac could do and what a comparable Windows laptop could do grew a little wider.

MacBook Neo Enters The Chat

By 2026, the M5 has no real competition at its performance level. That is what five years of compounding advantage looks like.

This is why when Apple launched a $599 laptop using the iPhone chip it beat every machine at that price point in the form of a complete package.

Just this month, Apple announced the MacBook Neo at $599. $499 for students.

That is the cheapest new Mac Apple has ever sold in its own stores. By a significant margin.

The CEO of ASUS said publicly that Neo’s pricing came as a shock to the entire PC industry.

That is not a throwaway quote. ASUS makes some of the best Windows laptops in the world.

Their CEO calling an Apple product a shock to the industry means the math did not add up in a way they expected. Something changed.

The thing that changed is the chip Apple chose to put inside it.

The iPhone Chip Inside a MacBook

The MacBook Neo does not run an M-series chip.

It runs the A18 Pro. The same chip Apple put inside the iPhone 16 Pro in late 2024.

When Apple announced this, the reaction in tech circles was immediate and skeptical. People questioned if a phone chip would be enough for a MacBook until the benchmarks came out.

It turns out the A18 Pro beats Intel’s Core Ultra 9 285HX - a flagship laptop processor that sits at the top of Intel’s lineup - by 9%.

It beats the Snapdragon X Elite by 16%. It outperforms AMD’s Ryzen 9 9950X3D by 3%. The only chips that score higher in single-core are Apple’s own M4 and M5.

NotebookCheck’s headline after testing it was blunt: “Apple toys with the competition.” lol.

Now, single-core is not the complete picture.

In multi-threaded performance (tasks that use multiple processor cores) simultaneously the A18 Pro still sits closer to M1-level performance .

The A18 Pro has six cores total, two performance and four efficiency. The M4 has ten.

That gap shows up when you are rendering long-form video, running complex code compilation, or doing anything that needs every core working at full capacity for a sustained period.

But here is the honest truth about who Neo is for.

Most people doing most things on a laptop are doing single-threaded work. Browsing. Writing. Email. Video calls. Watching content.Lightly editing photos.

Those tasks are generally handled by single-core performance, and the A18 Pro is remarkable at all of them.

Apple claims Neo is up to 50% faster for everyday browsing tasks than the bestselling Intel Core Ultra 5 laptop at a similar price.

Reviewers tested that claim and broadly confirmed the number is accurate.

The reason Apple could put the A18 Pro in a Mac and make it work and make it good comes down to manufacturing economics.

The A-series and M-series chips share the same performance cores and are built on the same TSMC 3-nanometer process. They are like cousins.

But the A18 Pro is produced at iPhone scale. Hundreds of millions of units per year. That manufacturing volume drives the cost per chip down dramatically compared to the M-series, which is produced in much smaller quantities for Mac.

That cost reduction is what made $599 possible.

A phone chip made it into a Mac not because Apple cut corners but because the phone chip is genuinely that good, and building it at phone scale made it cheap enough to put in the cheapest Mac Apple has ever sold.

What Is Actually Inside The MacBook Neo

Aluminum body, Not plastic. Aluminum!!

At $599, every Windows laptop you look at is plastic. The Neo is aluminum in four colors:

1. Silver

2. Blush

3. Citrus

4. Indigo

By the way, all of the colors above come with color-matched keyboards, which is a small detail but a nice one at this price.

13-inch Liquid Retina display at 2408 by 1506 resolution, 500 nits of brightness, support for one billion colors, full sRGB color gamut.

This just means that the Neo’s screen is noticeably sharper and brighter than anything else you will encounter at $599 in a Windows machine.

A18 Pro chip comes with a 16-core Neural Engine for running Apple Intelligence and other lightweight on-device AI features locally without sending your data to a server.

The base MacBook Neo comes with 8GB of unified RAM and 256GB storage. The 512GB model is a step-up which also adds Touch ID.

Two USB-C ports. One on the left side running at USB 3 speeds, which is fast. One on the right side running at USB 2 speeds, which is slow in 2026. The headphone jack moved to the right side. No MagSafe. No SD card slot. No Thunderbolt.

Similar to MacBook Air, the Neo has no fans. It has basically no moving parts making noise. The machine is completely silent.

Neo’s battery is also similar to MacBook Air’s with a capacity of 36.5Wh that can power the device for up to 16 hours as claimed by Apple.

Reviewer tests do come close to 12 to 14 hours under normal workloads. Basically, All-day without question.

One thing different from every other MacBook made in the last decade: the Neo uses a mechanical trackpad. Not Force Touch.

Force Touch is what simulates a click through haptic feedback under the trackpad that is used in other MacBooks.

Neo’s mechanical trackpad physically clicks when you press it. The mechanical trackpad on Neo actually moves. It is a cost reduction.

Reviewers noticed it but most said it is still significantly better than any Windows trackpad at this price tier, which is genuinely true.

Neo Versus Everything Else at $599

At $599, the Windows world mostly offers plastic body laptops that flex when you pick them up, displays that look washed out, trackpads that feel imprecise, and batteries that need a charger by early afternoon.

Tom’s Hardware called the Neo “a spectacular budget laptop that should shock the PC industry” and said it “never makes you feel like you’re getting a lesser machine in exchange for affordability.”

Where Neo wins every day in comparison to Windows laptops at similar price-point:

• A chip that is faster than anything Intel or AMD ships in a laptop at this price

• Battery life that genuinely gets through a full day without stress

• Aluminum build quality that does not creak or flex

• A display that is sharper and brighter

• A trackpad that is more accurate even when mechanical

• An Operating System optimized for the hardware

Where Windows wins: raw specs on paper. 16GB of RAM is common on Windows laptops at $599 versus Neo’s 8GB ceiling.

More ports in general. More storage in base configurations. And if your workflow depends on Windows-specific software, the Neo is simply not the answer regardless of how good the hardware is. Gaming is not a conversation on Neo at all.

The honest summary: for the person buying a laptop to write, study, browse, call, lightly edit photos, and do everything a modern connected person does day to day - Neo is a better machine than almost everything Windows ships at $599.

By the way, that is not a fan statement. The reviews and comparisons support it too.

Who This Laptop Is Actually For

OK, the MacBook is great but it is not a silver-bullet. Apple’s targeting a specific audience and it is worth being clear about.

Students are the most obvious answer and Apple knows it.

$499 education pricing puts a real aluminum Mac running full macOS within the range of people for whom $999 was never realistic.

Chromebooks dominate 93% of American K-12 schools right now. Apple is looking directly at that market and pricing accordingly.

Another target are the people already inside the Apple ecosystem.

If you have an iPhone, an iPad, an Apple Watch - the Neo makes all of those devices work together in ways that a Windows laptop simply cannot. iPhone Mirroring, AirDrop, Handoff, Universal Clipboard, Continuity Camera.

All of it works out of the box, immediately, without any setup beyond being signed in to the same Apple account.

First-time Mac buyers who have been curious for years but could not justify $999 as the entry price. Neo removes that barrier.

People doing what most people do on a laptop. Writing, studying, browsing, video calls, streaming, viewing and editing photos casually, managing documents.

The Neo handles all of that without hesitation and does it faster than any Windows laptop in its price range.

MKBHD made a fun point about the 8GB ceiling.

He said if you’re asking about the RAM ceiling on MacBook Neo the laptop is straight-up not for you :p

Because if you know 8GB is low for you, your workflow probably needs more and the architecture of MacBook Neo doesn’t allow RAM to be upgraded. So, that makes it a bad choice for such people.

If your work genuinely requires more than 8GB and some people do like developers, video editors, the Neo is not your machine and you should know that before you buy it.

The MacBook Air at $1,099 starts with 16GB and handles those workloads more comfortably.

Who the MacBook Neo is not for:

• Developers running multiple environments or virtual machines simultaneously

• Designers who need P3 wide color gamut coverage (Neo covers sRGB, not Display P3),

• Anyone whose software runs on Windows and has no Mac equivalent

• Anyone who games seriously

Know your needs clearly before you buy. Watch reviews online and see benchmarks for your specific workflows like how the apps run on the MacBook Neo as reported by people online.

The Neo is excellent at what it is designed for but genuinely limited at everything outside of that.

The Most Repairable MacBook in 14 Years

iFixit is the gold standard for repairability scoring. They tear down every major consumer device, document every component, and score it on how easy it is to repair if something breaks.

Apple products have historically scored terribly on their scale. The 2012 Retina MacBook Pro earned a 1 out of 10. Recent MacBooks have sat between 3 and 5.

When iFixit tore down the MacBook Neo, they said something they have not said about an Apple laptop in a very long time. They said it sent cheers across the office.

The battery - instead of being glued to the chassis with industrial-strength adhesive like every recent MacBook before it - is secured with 18 screws into a tray.

Screws can be removed with a screwdriver. Adhesive cannot be undone without heat guns, specialized tools, careful prying, and the real risk of puncturing the actual battery.

Battery replacement on recent MacBooks required all of that and cost $129 to $199 at an Apple Store. On the Neo, it is a straightforward job.

The USB-C ports are modular too. The headphone jack is modular. Both are color-matched to the chassis.

If a port breaks - which happens a lot, especially on laptops owned by students who plug and unplug things constantly it is a port replacement, not a logic board replacement.

That distinction is enormous. A logic board replacement on a MacBook costs hundreds of dollars. A modular port replacement costs a fraction of that (if available for fair price though).

The keyboard can be replaced without swapping the entire top case, which is how previous MacBook keyboard repairs were handled and why they cost so much.

Apple published a full repair manual for the Neo on the same day it went on sale. Not a few months later. Day one.

iFixit gave it a 6 out of 10. The M5 MacBook Pro scores 4 out of 10. The M4 MacBook Air scores 5. A score of 6 from iFixit is the highest any MacBook has scored in approximately fourteen years.

iFixit’s exact words: “We haven’t been as happy about a MacBook since 2012.”

This did not happen by accident. Apple is targeting the education market and school IT departments care enormously about repairability.

A broken Chromebook in a school can be fixed quickly and cheaply because the parts are accessible and the documentation is public. Apple designed the Neo to compete in that world. The repairability is not a side effect of the design. It is the design.

The Tradeoffs You Should Know

Let’s start with the obvious ones.

No backlit keyboard in 2026 is genuinely difficult to defend. Every Windows laptop at $599 has backlighting. Every other MacBook has backlighting. Neo does not.

Apple says the white keycaps are more visible in low light than darker keys. They are, slightly. It is not enough. Typing in a dim room, on a flight, or in bed without disturbing someone is a real use case and the Neo makes it harder than it should be at this price.

This is the most fair criticism of the machine and it is completely valid.

8GB of unified memory is soldered directly to the logic board. There is no upgrade path. The machine you buy today is the machine you will have in five years.

With Apple Intelligence being memory-hungry and the general direction of software toward requiring more RAM over time, buying a machine with an 8GB ceiling is a decision worth thinking about carefully.

If there is any chance you will push this machine professionally, buy the Air with 16GB.

The right-side USB-C port runs at USB 2.0 speeds. 480 megabits per second. External SSDs, modern USB accessories, and high-speed docks can transfer data at thousands of megabits per second.

Plugging a fast drive into the wrong port on the Neo and wondering why transfers are slow is a trap that is easy to fall into.

External display support tops out at a single 4K 60Hz monitor. One screen, that is your maximum. Multi-monitor setups are not an option with MacBook Neo.

The speakers are noticeably weaker than other MacBooks. Multiple reviewers flagged this consistently and independently. The side-firing speakers do the job for casual listening but they are clearly the place inside the chassis where the $599 price point shows most visibly.

None of these are dealbreakers for the person the Neo is designed for. But they are real tradeoffs and you guys should know about them before spending your money.

Price, Performance & Package

Here at SK NEXUS, we like to define the Apple Silicon MacBooks by three Ps, those are; Price, Performance and Package.

It is a simplified description of how we view MacBooks that is by Price, Performance and the overall product as a package.

Compare a similarly-priced Windows laptop from any of the big laptop manufacturers like HP, Dell, Lenovo with Apple Silicon MacBooks and you’d see their price; they don’t deliver near as much for the price of a starting MacBook.

When you compare the Apple Silicon MacBooks released in 2020, their performance is unmatched when compared to Windows laptops in the same price bracket.

And then there are overall components like the screen, speakers, touchpad, keyboard, full aluminium build, MacOS, all of that make up the final package that is the MacBook.

It is 2026. The M5 has no meaningful competition at its performance level - Intel, AMD, and Qualcomm are all chasing a target that keeps moving further away.

The MacBook Air has been the best laptop in terms of Price, Performance and the overall Package that most people can buy for three consecutive years.

And now the Neo is arguably the best laptop at $599. Apple owns nearly every tier of the laptop market simultaneously. That has never been true before in the history of personal computing.

The path from the butterfly keyboard disaster of 2015 to this specific moment runs directly through one decision made in 2020: ditching Intel and building their own silicon.

That decision took five years to fully show itself.

Year one, M1 shocked everyone. Year two, M2 refined what M1 started. Year three, M3 added efficiency the previous generation could not match. Year four, M4 pushed the performance ceiling further than any competitor had matched.

Year five, M5 made the gap so wide that even the cheapest Mac Apple sells can use a chip from last year’s iPhone and still beat everything Windows ships in its price range.

Every year the MacBook Air kept winning its category. Every year more reviewers switched their recommendation. Every year the comparison articles between Mac and Windows get harder to write in Windows’ favor for the majority of users.

And then Linus made his video recommending MacBooks only.

Linus switching is not just a meme. It is a signal.

When the person whose audience is built on PC building and Windows enthusiasm looks at what Apple has done with silicon and says he cannot recommend anything else in good conscience, that reflects a genuine shift in what is true about these products.

He did not switch because he likes Apple. He switched because the hardware became impossible to argue against for most of his audience’s needs.

The ASUS CEO calling Neo’s pricing a shock to the industry is another signal. He isn’t just a regular executive. He publicly said that the PC industry did not see it coming.

And honestly, no-one did.

The question now is not whether MacBooks are good. Everyone agrees they are. The question is whether anything in the Windows ecosystem is going to meaningfully close the gap, and when.

Competition is coming. Maybe it closes the gap. Maybe it does not but as things stand, MacBooks are winning.

In March 2026, Apple built a $599 laptop with an aluminum body, an all-day battery, a chip that beats every Intel and AMD competitor at its price tier, and the most repairable design any MacBook has had in fourteen years.

That is not an ordinary product launch and that’s why everyone’s talking about it.

My Thoughts on MacBook’s Impact

Apple has been a trailblazer in adopting the ARM-based processors for general purpose use and it has paid them dividends.

To me it feels like MacBook may just win this decade in consumer laptops in price, performance and package.

It’s 2026 already and we don’t have any serious competition to MacBooks. Right from the start of this decade Apple has kept the laptop crown.

And 6 years later, it is still maintaining its lead. But it’s not like the companies aren’t trying. In fact, the competition has started to heat up.

Nvidia is rumored to enter the laptop CPU space with ARM-based chips later this year. Qualcomm is improving its Snapdragon X chips which have performed great.

That Qualcomm Snapdragon chips are already being used and have outstanding reviews especially when compared to x86 mobile chips.

Qualcomm is trying their best to compete with Apple Silicon even when Apple has a vertical integration edge over its laptops.

Vertical Integration basically means that Apple has complete control over the hardware as well as software stack of their MacBooks and that allows them to push optimizations that other laptop manufacturers just could not.

Windows has been one of the biggest bottlenecks in the PC industry for quite some years now and Microsoft is also promising to fix it from the ground-up.

They’re already maintaining ARM-builds for Windows and re-writing core Windows libraries in modern languages like Rust.

With time the Snapdragon chips will get mature and as that hardware ages a couple of years it will land the used lot market in Pakistan.

That is when we’ll go from good-for-most laptops that easily last a day that would be available for cheap prices here.

For me, I’ll be happily recommending Macbook Neo and similar price-point laptops especially when they land inside our used market.

And while Apple’s busy beating the hell out of every other processor, It’s not like the CPU wars in the PC space are over. Intel just launched its Core Ultra 200K Plus series and reviewers are calling it a genuine comeback.

This time we not only have x86 vs ARM but x86 is also redeveloping itself against the performance that ARM-based processors have delivered.

Stay tuned for a whole article discussing what Intel’s been doing lately and what impact could its latest chips have for Windows desktop and laptops moving forward.

Comment down if you’re excited for my article on Intel Core Ultra 200k series.

Further Learning

• Apple toys with the competition - MacBook Neo

• MacBook Neo Is the Most Repairable MacBook in 14 Years

• Apple MacBook Neo review: a budget-priced game-changer

• Budget speed: How fast will the low-cost MacBook Neo really be?

• Apple has officially ended its butterfly keyboard

• Apple Abandons Its Breakable Butterfly Keyboard for Good

• I Can Only Recommend Macbooks Now…

• DAMN, Apple - MacBook Neo

• Apple is Terrifying the PC Industry

• Macbook Neo Review: Better than you Think!

• Mac vs Windows - It’s not close in 2026!

• The 2018 MacBook Pro has a problem with overheating

• MacBook Pros Are Underperforming. A Software Fix Is Coming

• i9 Macbook Pro 2018: Hottest Laptop on the Planet!

• Review: Apple MacBook Neo

Looks like you made it to Part 3 and I respect the consistency.

A quick recap - In the previous piece, we focused on communication aka choosing the right channel, understanding encryption, and accepting the tradeoffs that come with it. If you missed that one, go back and read it first. This builds directly on it.

For anyone new here: this is Part 3 of a practical privacy series designed to help you secure your digital life step by step.

Now here’s a little surprise - even if you’ve done everything right so far - hardened authentication, improved your communication tools - you’re still exposed.

Why?

Say hello to your browser.

Most of your digital life flows through it. Whether it’s your email, work dashboards, banking. Social media - Your browser quite literally sees everything. And even with strong logins and encrypted messaging, browsers still collect and leak behavioral data in ways most people never think about.

But don’t worry - we’re going to fix that.

And before starting, just a quick heads up for the new folks, If you’re unsure about the sequence, follow this order:

• Privacy Is Not Security - The Difference is Important

• Privacy Fundamentals - Part 1: Authentication

• Privacy Fundamentals - Part 2: Communication

How Web Tracking Actually Works

Every website you visit has ways of remembering you. Some of those ways feel harmless while some of them don’t.

The most common method is the cookie. A cookie is a small text file your browser stores when you visit a site. It keeps track of things like your previous visits, preferences, and activity. That’s why your shopping cart doesn’t reset every time you refresh the page. It’s why websites remember your language settings or keep you logged in.

There are two kinds worth knowing about:

First-Party Cookies

First-party cookies are created by the site you’re actually visiting. They’re part of the agreement you implicitly make with that site: “Remember my preferences so this works better.” If you add items to a cart or log into an account, that memory lives inside a first-party cookie. The information stays with that domain.

Third-party cookies

These are set by a completely different domain than the one in your address bar. Usually advertisers, data brokers or analytics firms. Third-party cookies are mainly used for three things:

• Tracking - They track you across different websites to see what you may buy.

• Retargeting - They send you to a website that sells products you may like.

• Ad-Serving - They deliver personal ads that target your desires.

Btw these cookies are almost everywhere - one study found third-party tracking cookies on 92% of websites.

Anyway, let’s take the following example to understand third-party cookies better.

You visit a news site, but behind the scenes it loads content from ad networks. Those networks drop their own cookies into your browser. If that same ad network appears on thousands of websites, it can recognize you on every single one of them through the tiny identifier stored in your browser aka cookie.

This is called cross-site tracking and cookies are just one of many ways to achieve cross-site tracking.

How Tracking Pixels Work

Websites also embed tracking pixels. Pixels are tiny invisible 1×1 images or scripts. You’ll never see them. But when your browser loads the page, it automatically contacts another server. That server logs your IP address, your device type, your browser, the page you just visited and much more. If there’s already a third-party cookie in place, it connects the dots.

So while you think you’re just reading an article, multiple companies may be notified that you were there.

For example, if a page includes a tracking pixel from a large social media company, your browser reports your visit back to that company the moment the page loads.

You can put it this way - a pixel doesn’t identify you by itself. Instead, it acts as the messenger while cookies hold all the actual tiny details about you.

Together, they create something bigger - “cross-site tracking” aka the ability to follow the same person across completely different websites. And cross-site tracking can be achieved through various other methods such as: Browser fingerprinting, Cname cloaking and many more techniques (I won’t be talking about all of them.)

Tracking Through Fingerprinting

A digital fingerprint is exactly what it sounds like - a list of characteristics that, when combined, point to a single device.

Every time you open a website, your browser sends information so the page can load properly because it has to. Things like your browser type, operating system, language settings, and approximate location. But that’s just the beginning. Scripts running in the background can also see your screen resolution, time zone, installed fonts, device memory, graphics settings, and dozens of other small technical details.

At first glance, the pieces feel mundane, but when compiled and analyzed at scale, they create a stable identifier. One that can follow you across websites.

Tracking companies stitch these fragments into a unique profile that distinguishes your device from millions of others by just observing how your system is configured.

It’s important to note that Ad networks rely on two primary ways to keep tabs on users: cookie tracking (which we’ve already discussed) and browser fingerprinting. Cookies store an ID inside your browser. Fingerprinting doesn’t need to store anything at all. It simply recognizes you when you show up.

And because most devices are typically used by one person, identifying the device often means identifying the user.

In short, everything observable gets collected. Everything collected gets compared and the result is a persistent identity tied to your browsing behavior.

For anyone wondering whether clearing cookies solves the problem - this is why it doesn’t. Browser fingerprinting bypasses that reset button entirely. And if you want to test out how much data is collected when you visit a website - check out this website, and click “test your browser.”

Chrome - The Default Browser With a Business Model

Most users just open their browsers and start using it. For the majority of the world, that browser is Chrome. Recent market share estimates consistently show Chrome controlling well over 60% of global browser usage. In other words, when we talk about how the web works, we’re largely talking about how Chrome works.

Truth be told - Google is one of the most powerful advertising machines ever built. While people see Google as a search engine, its core business is advertising. The overwhelming share of its revenue comes from selling targeted ads.

And targeted ads require data.

Chrome and Google Search are two of the most efficient data collection tools ever deployed at scale. Every search query and interaction feeds into a system designed to refine what it knows about you. The more signals Google gathers, the sharper its ad targeting becomes. The sharper the targeting, the higher the ad revenue.

Open Chrome right now. Look at the top right corner. If you see your profile picture or your name, you’re signed in. That means your browsing activity may be linked directly to your Google account.

Don’t remember signing in? You’re not alone. Chrome has increasingly nudged and in some cases automatically signed in - users who log into Gmail or other Google services.

On mobile, it goes further. If you’re using Android, Chrome is tightly integrated with the operating system. Location data can be transmitted when you search. Even when users disable location sharing, less precise coordinates can still be inferred.

If you want to understand just how central Chrome is to the modern tracking ecosystem, deeper investigations are worth reading - particularly reporting from The Inquirer and independent privacy researchers who have examined Chrome’s role in surveillance advertising.

Privacy-Focused Browser Options

There’s no shortage of browsers in the market. In recent years, newer privacy-centric projects like Mullvad Browser and Kagi’s emerging browser have emerged.

That said, I’m only recommending tools I’ve personally used and tested.

Firefox

Mozilla Firefox remains one of the strongest all-around browsers for privacy and security. It combines solid built-in protections and a track record that’s difficult to ignore.

It stands as a genuine alternative to Google Chrome, Microsoft Edge, and even Safari.

There’s a reason Tor Browser is built on Firefox. The Tor developers chose it because Firefox can be hardened and locked down in ways Chromium-based browsers simply cannot match. It’s also why you’ll see Firefox set as the default browser across many Linux distributions.

Put simply - Firefox has earned its reputation.

That said, out of the box, it’s not fully optimized for privacy. You’ll need to spend some time hardening it.

A clear, step-by-step guide is available at Privacy Guides. Navigate to “Desktop Browsers,” scroll to Firefox, and you’ll find detailed instructions on which settings to enable or disable:

And if documentation isn’t your style, there’s also a practical walkthrough video available here:

Brave

Brave and Chrome are both built on Chromium, the open-source browser engine that also powers Edge, Opera, and Vivaldi. On the surface, they’ll feel familiar and any Chrome extension will work on Brave.

But the real difference is under the hood.

Brave ships with many privacy protections enabled by default. In Chrome, achieving similar functionality would require installing multiple extensions - each one adding overhead and potential security risk.

In practical terms, Brave delivers stronger default privacy with less configuration.

If you want a deeper comparison between the two, these resources break it down thoroughly:

• https://www.droidviews.com/brave-browser-vs-google-chrome/

• https://brave.com/compare/chrome-vs-brave/

You can download brave by visiting this link and for maximizing privacy settings, this configuration guide walks through it step by step:

One important thing to note - switch your default search engine from Google to Brave Search, Qwant, or DuckDuckGo. All three are privacy-respecting options. It ultimately comes down to preference. Lastly, don’t overthink about which browser to go with. Either Firefox or Brave, when properly configured, will significantly reduce passive tracking compared to Chrome.

For transparency - I’m currently using Brave, I prefer the interface.

Extra Protection Tools

The following extensions are optional. But if you want to push your setup further, they’re worth considering.

Ghostery

You can Install Ghostery directly by visiting this link.

Out of the box, its default configuration is solid. You don’t need to tweak much unless you want to.

Ghostery acts as an additional filtering layer on top of your browser’s native protections. It strengthens tracker blocking and improves visibility into who’s attempting to collect data on you.

Password Manager Extension

Although this isn’t directly about tracking, it ties back to Part 1 of this series.

If you’re using a password manager (which you should be) install its browser extension. Visiting the vault website every time adds friction. Friction reduces consistency. Reduced consistency weakens security.

I personally use Bitwarden. If you do as well, you can get the extension here

Optimizing Your Behavior to Build Privacy Habits

keep in mind, neither any browser nor any hardened configuration can protect you if your own behavior keeps undoing the work. At some point, you are the control surface and if you’re careless, all your efforts will be for nothing.

Privacy is alignment between what you want and how you operate daily. Without that alignment, even the strongest setup erodes over time. And if you want a private digital life, your daily behavior has to match it.

Separate to Stay Separate

One of the simplest and most effective privacy habits is isolation.

Use separate browsers or at minimum, separate browser profiles - for different parts of your life. For example: you can use one browser profile for work and the other for personal browsing. What’s important is to - keep them isolated.

Why?

Because correlation is how companies map you.

When your banking sessions, social media activity, shopping behavior, and professional research all flow through the same browser identity, it becomes child’s play for companies to stitch together a detailed profile about you.

But, segmentation can disrupt that.

What’s interesting is, you can take this same logic and scale it up.

One email for banking and formal communication.
One email for social media.
One junk email for online shopping and signups.

You can take it further if you want.

One device per activity.
One network per activity.

At that point, you should ask yourself “How private do you actually want to be?” Then use this logic and adjust things accordingly.

Permissions, Extensions, and Self-Inflicted Leaks

Earlier, I mentioned browser extensions. They are powerful, but that power cuts both ways. A poorly vetted extension can compromise your entire environment and this is more common than you think.

The first problem involves many users installing extensions casually and not even verifying if these extensions are safe to use.

Excess permissions are another problem: clipboard access, microphone access, camera access, location access. If a website or extension doesn’t absolutely need a specific permission, don’t grant it.

To understand how dangerous malicious extensions can be, consider a real-world case.

A malicious extension posing as a ChatGPT search tool remained available for months. Some compromised extensions had behaved legitimately for years before being activated. Once triggered, these extensions deployed mechanisms that intercepted user traffic.

Malwarebytes researchers described a scenario like this:

“You receive a Zoom meeting invitation and click the link. Instead of joining your meeting, one of the malicious extensions intercepts your request and redirects you to a convincing fake page claiming you need to download a “critical Zoom update” to join. You download what appears to be legitimate software, but you’ve just installed additional malware onto your system, potentially leading to full machine takeover and complete compromise of your device.”

Anyway, what you can do right now is identify the different extensions installed in your browsers, ask if you genuinely need them? Ask if they are actively maintained and reputable?

If the answer is “no” - remove it.

And if you suspect compromise:

• Remove the extension.

• Change your passwords everywhere.

• Reinstall your operating system clean.

The Future of Browsing

We’re entering a new phase.

AI-native browsers are already appearing - OpenAI’s Atlas, Perplexity’s Comet, and others positioning themselves as the future of web interaction. All of them pitching the same thing - smarter browsing, seamless automation, personalized assistance…etc.

And increasingly of course - better privacy.

I do believe that this category will grow and AI browsers will become mainstream products.

But many of these platforms market privacy as a feature while operating on data-intensive models. The more intelligent the browser claims to be, the more context it needs to function effectively.

On top of that, novelty often comes at the expense of security maturity.

Comet being exploited was a reminder. New platforms frequently ship fast and harden later making early adopters mere scape goats.

So the advice remains simple - don’t get distracted by marketing, don’t confuse AI features with privacy guarantees and just stick to the fundamentals.

In the next piece, we move one layer deeper - beyond the browser itself. We’ll look at the network layer: networks, tunnels, and much more. That’s the final layer in this privacy series. The last cherry on top.

Stay tuned.

Thanks for reading.
See you next time.

The Algorithm is Winning!

I got a confession - most of my time goes into researching on topics - so, I can help you see what the system doesn’t want you to see. At this point, it’s me versus The Algorithm and the Algorithm is winning…

That’s why I need your help.

If SK NEXUS has earned a place in your inbox, consider subscribing if you haven’t already.

And if you’re active elsewhere, follow along on Bluesky or X (Twitter) or LinkedIn - those are three two places I show up most consistently outside Substack. A follow, a like, a repost - it matters to me, more than you think.

Independent research only works if the signal travels. If this work has been useful to you, help it travel.

What You Might Like

It’s been some time since IoT news started making headlines again, especially in the sea of anything AI.

Of course we just can’t get enough of discussing AI models these days, New iPhones are always in-trend and gamers are also pretty vocal about kidney-priced GPUs.

But a few months ago, something quieter happened in the hardware scene.

Qualcomm - the smartphone chip behemoth, the company behind many of the processors inside our phones - acquired Arduino, the tiny board that taught millions how to build electronics.

Arduino is the reason thousands of students first touched hardware.

It’s the reason hobbyists built smart homes before “smart home” was a thing. It’s the reason countless startups could prototype products without a mountain of cash.

At first glance, it sounds niche esp if you’re not into DIY electronics but let me share with you that it directly impacts some odd 30 million IoT developers, and not just them but the future of hardware in classrooms, startups and inside our homes.

Arduino’s acquisition isn’t your normal corporate transaction. It isn’t just about a big company buying a smaller startup. It’s about the future of accessible electronic hardware.

This article is an in-depth look at the acquisition coming from someone who owns and loves tinkering with his Arduino board.

What Is Arduino?

If you’ve never heard of Arduino, think of it like this:

It’s a palm-sized circuit board with a tiny microcontroller on it. You plug it into your laptop, write a few lines of code, and suddenly you can control lights, motors, sensors - basically physical stuff.

As my friend ChatGPT phrases it: Arduino is the Lego set of electronics.

So, Arduino is a microcontroller board with the ability to digitally connect and control other hardware connected to it. It could be easily programmed just through a USB cable.

Inside an Arduino board is something called a microcontroller. That sounds complicated, but it’s just a tiny computer designed to run very small but fast programs.

Most of your ‘smart’ electronics have some kind of a microcontroller baked-in. Your microwave has one. Your washing machine has one. Traffic lights use them. Cars use many microcontrollers to automate different tasks.

Microcontrollers are generally used to automate one small task like turning off a motor, controlling a switch remotely, etc. They’ve historically been accessible to large companies or specific industries.

What Arduino did was take that tiny computer that engineers used in factories and make it usable by normal people. No advanced lab. No corporate license. Just USB cable, simple software, and your imagination.

These days Arduino is more of an ecosystem now that is not just limited to the physical dev-boards but also includes their own programming software (Arduino IDE) and programming language (Arduino C).

Arduino is known for popularizing DIY electronics by making microcontrollers accessible outside just professional engineering circles.

Note: It’s easy to mix an Arduino with a Raspberry PI.

Even though both share many features and may look similar on surface, they are different devices. You may check our detailed article below on Single Board Computers to learn more:

Hardware scene before Arduino

Before Arduino showed up in 2005, programmable hardware wasn’t as accessible or friendly for normies. Specialized boards existed but they were expensive and locked down too.

Moreover, If you wanted to program a microcontroller, you had to use complicated programming languages (aka low-level languages) C or C++.

You also had to deal with vendor-specific software for each hardware and sometimes buy separate programming hardware.

Before Arduino, we had specialized boards that were not only hard to get a hold of by a normie but were expensive too. Back when the founders of Arduino started, some development boards would cost more than $100.

It wasn’t impossible but it wasn’t welcoming for our average Bashir. Nerds could go through these lengths but most high school students don’t want to.

Most development boards before Arduino were proprietary. Documentation was hard to find and communities were small.

In simple terms: hardware was for trained engineers not curious individuals sitting in their homes who want to experiment with their DIY skills.

The Open Source DNA

Arduino was created in-response to limitations in microcontroller boards of the time. The idea was simple: to create a more open and accessible microcontroller board that is easy to program for students.

One of the co-founders of Arduino was Massimo Banzi who was tasked to teach students computer electronics.

Early on, Banzi discovered that microcontroller boards of the time would cost about $100, didn’t have computing power for complex programs and weren’t compatible with the popular Macintosh computers at the time.

Banzi was lucky to have a colleague at MIT who had designed a simple programming language for new programmers. The better thing was that the project was open-source. That was the inspiration for Arduino IDE.

Note: IDE stands for Integrated Development Environment. In our case, an IDE is just a collection of specialized software that allows one to write and push their programs onto the microcontroller.

A student in Banzi’s program started working on a related project called Wiring, which had a user-friendly IDE and a simple circuit board.

The important thing here is that Banzi and his peers were strong believers in open-source software.

Because their goal was to create a quick and easily accessible platform and they feared the institute won’t be able to fund the project, the co-founders felt they’d be better off opening up the project instead of keeping it closed.

Here’s how one of the co-founders recalls their decision to go open:

“So we said, ‘Forget it,’” Banzi recalls. “ ‘Let’s make it open source.’”

How Arduino Hugged Open-Source

The initial Arduino software and more importantly - the hardware design itself was free and open-source for anyone to use and collaborate on. There was no complex licensing, no nothing!

The initial discussions around pricing the board wanted it to be priced around $30 or equal to going out to dinner at a pizza place. Right from the start, they wanted it to be student-friendly.

You plugged the board into your computer via USB. You opened the Arduino IDE. You clicked “Upload.” That was it.

It also shipped with example code. Want to blink an LED? There’s a template. Want to read a temperature sensor? There’s a template.

That meant people could study it, modify it, even manufacture compatible versions.

Tech publications have highlighted how Arduino’s open hardware model made it popular globally by lowering costs and encouraging clones.

Arduino’s Explosive Impact

Arduino solved the problem of making accessible boards but it didn’t stop there and instead built an ecosystem that combined that hardware with their software.

Their result was an affordable kit that was simple to program and play with even for a high school student.

Because the Arduino designs were open, other companies could make Arduino-compatible boards. These clones often cost one-third the price.

That meant students in developing countries like ours could afford them. Schools could buy them in bulk. Makerspaces could stock them cheaply.

Arduino’s community has also been central to its open identity. Makerspaces around the world love Arduino and there are thousands of projects built on Arduino that are freely available on the internet.

This sense of community work has also contributed to Arduino’s success. For a newcomer much of the work is already done by the community.

The Arduino community has a plethora of software libraries (pre-written programs) that beginners could just plug and play in their projects.

Impact on Education

Arduino’s plug and play experience thanks to the community, is how you see high school students walking around with their prototype robots being controlled over an Arduino board.

Instead of static programs printing “Hello World” on a monitor, students could have their small programs control LED lights, steer their robots or fly their drones.

The Arduino community not only reduced the bar to experience Arduino but also made it fun for beginners.

In a way, getting into engineering and Computer Science courses was fun thanks to Arduino’s community that has their projects freely available on the internet.

This community and plug-and-play experience is why it’s hard to see an Electrical Engineering or a Computer Science without mention of an Arduino board.

Arduino became the gateway into engineering.

That’s why tech outlets have repeatedly described Arduino as foundational to the modern maker movement and STEM education growth.

For many startup founders today, Arduino was their first hands-on experience with how software controls hardware.

Impact on Small Businesses

For businesses, Arduino wasn’t usually the final product inside your house but for many small businesses it was often the starting point.

Startups used Arduino to prototype hardware cheaply. Instead of designing custom PCBs from day one, they tested ideas using Arduino boards.

Many home automation setups, environmental sensors and irrigation related startups played around with Arduino boards before finalizing their products.

Industry analysts have noted that Arduino dramatically reduced early-stage hardware development costs for small teams and startups.

In a sense, Arduino made it easy for hardware engineers to experiment and build and test early versions of their product.

It reduced costs and removed the friction one used to face before Arduinos were a thing.

So Why Did Qualcomm Buy Arduino?

Whether we like it or not, it’s a fact that every device wants to connect to your WiFi and call itself ‘smart’.

Smart devices are everywhere. Edge AI is rising. Even basic products now ship with sensors and connectivity.

Qualcomm wants to be central to that future and Arduino sits at the entry point of that future.

That’s why just last year in October, news came out that Qualcomm just acquired Arduino (the company).

We all know that Qualcomm isn’t a hobby brand. It designs smartphone processors which are way more powerful and complicated than an Arduino.

The company announced it was acquiring Arduino as part of a broader push into AI and IoT devices. From Qualcomm’s perspective, this makes strategic sense.

Arduino has millions of developers and students using its platform. They can influence them in earlier stages to be more familiar with their stack.

If those builders eventually build companies, whose chips would they be more familiar with?

Industry analysts say that this acquisition is Qualcomm trying to “meet developers earlier in the pipeline” of hardware innovation. If you didn’t get it, here’s a meme:

What Changes Technically?

One early sign of direction was the announcement of the Arduino UNO Q, a board combining Arduino compatibility with Qualcomm silicon and more advanced capabilities like Linux and AI support.

The Verge reported that this board aims to blend Arduino’s beginner-friendly ecosystem with more powerful computing hardware.

That means future boards could include:

Better wireless connectivity
Built-in AI acceleration
Stronger integration with modern IoT stacks

For users, that sounds like progress.

But progress always comes with trade-offs.

The Openness Question

Here’s something that worries the open-source community.

Arduino built its brand on openness, free designs and community. Qualcomm built its empire on proprietary silicon and licensing.

So their concern makes sense:

Will Arduino remain open?

Reuters reported that Qualcomm said Arduino would continue supporting multiple chip vendors and maintain its mission.

On paper that sounds good, but corporations don’t think twice before contradicting their own statements.

And open-source communities generally don’t trust big companies nor their press releases. Trust is the most essential element to these small communities and that’s why they’re skeptical.

If openness shrinks, clone compatibility disappears, or licensing tightens, the ecosystem could fragment.

And once a maker ecosystem fractures, it rarely fully recovers.

What This Means For Students and Hobbyists?

At the moment we don’t have much to worry about. If you’re a student, here’s the short version:

In the near term, nothing dramatic changes. Arduino boards still work. Tutorials still exist. Their Arduino IDE still works as it should.

In the medium term, we can expect more powerful boards with built-in connectivity and AI features because Qualcomm already has a proven portfolio in that sector.

That could actually be exciting for us Arduino fans. But affordability and openness will determine whether Arduino stays true to its original mission or not.

For hobbyists and tinkerers like me, this acquisition could mean:

More powerful toys which means more complex projects.

For startups, it might mean smoother scaling from prototype to commercial silicon - especially if one wants to integrate with Qualcomm’s tech.

My Somewhat Optimistic Take

Arduino’s story is personal to me because Arduino UNO was one of my first interactions with a microcontroller board.

On a random night, I had a project idea related to RFID badges, I still remember going through it in detail with ChatGPT which had just come out.

The very next day I was able to buy an Arduino clone for ~900 PKR ($3) from my local electronics market and my project was ready that week.

And that was my first experience with Arduino. I didn’t have to find the schematics of the microcontroller or write Assembly language code just to build out my small project.

That heavy-lifting was already done by Arduino and the community. It was largely a plug and play experience for me.

The more important thing here is that because of Arduino’s open design we (in developing countries) can get Arduino clones for that cheap.

Arduino encourages others to use their designs. An original Arduino that costs me 9,000 PKR ($30) is available for 900 PKR ($3), and for me, both of them are identical in feature-set and performance.

That’s the power of that open-design philosophy.

It’s a fact that without Arduino, hardware would still feel scary to millions and engineering courses wouldn’t be that fun.

It’s a reality too that without companies like Qualcomm, many connected devices wouldn’t exist at scale. The company has shipped like hundreds of millions of processors, just let that sink in.

Arduino’s acquisition connects the biggest open hardware ecosystem with a technology behemoth.

If Qualcomm preserves Arduino’s open spirit, keeps boards affordable, and strengthens and not restricts the ecosystem, this could be an evolution.

If not, it risks becoming another story about corporate greed and a failure of capitalism.

For now, the DIY/maker movement just got a new owner. The next few years will show us which path this takes

And people like me who ever plugged in that little blue board would be watching closely.

As always thanks for taking out the time to go through my blogs. I appreciate all of my readers. I’d love to know about your thoughts down below. Your comments actually make my day.

Any feedback or criticism is welcome down below in the comments!

Further Learning

• The Making of Arduino - IEEE Spectrum

• Qualcomm Buys Arduino, and the Open-Source Community Is Skeptical

• How Arduino changed the world

• How does Arduino detonate the world of open-source hardware makers?

• Arduino and the Open Hardware Revolution

• Qualcomm acquires Arduino in surprising move

• Behind Qualcomm’s Arduino Acquisition: 33 Million IoT Developers

• 15 ingenious projects you can build with an Arduino - XDA

If you’re here, you’re trying to take control of your digital security and that is a good thing.

In the previous piece, we focused on authentication - locking down access, strengthening credentials, and reducing account takeover risk. If you missed that one, go back and read it first. This builds directly on it.

For anyone new - this is the Part 2 of a practical privacy series designed to help you secure your digital life step by step.

Part 1 was about who gets in.
Part 2 is about what goes out.

A lot of people harden their passwords, enable two-factor authentication - and then discuss sensitive matters over insecure channels every day which kills the whole point of privacy and securing yourself.

But worry not, cause I am here, I’ll be guiding you on how to patch this bug in your life. And before starting, just a quick heads up for the new folks, If you’re unsure about the sequence, follow this order:

• Privacy Is Not Security - The Difference is Important

• Privacy Fundamentals - Part 1: Authentication

How Digital Communication Actually Works

A lot of people think of digital communication as a direct exchange - one device sending a message straight to another.

But, that’s not how it really works.

Messages normally pass through multiple layers before reaching their destination. They move through servers operated by service providers. They may be stored temporarily or permanently on cloud infrastructure. Logs may record when messages were sent, from which device, and to whom.

Even when a message appears instant and private, it is usually traveling through systems designed for reliability and data analysis.

It’s important to note that modern communication platforms prioritize convenience, and analytics. Privacy is not their default objective.

What End-to-End Encryption Means

You must have seen this term tossed around here and there, but what does it mean?

When we talk about end-to-end encryption (E2EE), we mean something very specific. Only the sender and the receiver can decrypt the message.

In simple terms, only you and the person you are communicating with hold the keys needed to read the content. Neither the platform, nor anyone else should theoretically be able to view the content of the messages.

That means the message content - what you actually wrote - cannot be read by anyone else while it is in transit. If you want a deeper technical explanation of how E2EE works, including how encryption keys are generated and exchanged, you can read this article:

Now, you know what E2EE is, but another important thing to understand is what it protects and what it doesn’t protect.

What E2EE Protects:

• The content of your messages

• Attachments you send

• Voice notes or media shared within the encrypted channel

What E2EE Doesn’t Protect:

• Metadata (who you contacted)

• Timing (when you communicated)

• Frequency of communication

• Contact lists…etc

So, yes - even with the use of encryption, service providers may still see who is talking to whom and when. Encryption only protects the contents of the message. It does not automatically protect the surrounding data.

Why Encrypted Doesn’t Mean Private by Default

There’s a common misconception that if an app claims to use encryption, it must automatically be private. That’s not how it works. I even talked on this in detail in one of my previous posts, about how security doesn’t guarantee privacy, if you haven’t read that till now, check it out:

As I mentioned earlier, End-to-end encryption protects the content of messages while they are being transmitted. It prevents outsiders from reading what was sent. But encryption alone does not determine how a company operates, what data it collects, or how it handles user information.

An app can advertise encryption while still collecting metadata. Backups may not be encrypted. Servers may retain logs. The user’s personal device itself might just be compromised. None of these risks disappear just because messages are encrypted in transit.

It’s also worth noting that on some social media platforms, end-to-end encryption is not enabled by default. Instagram is a good example of this. E2E encryption is not automatically turned on for chats, you have to enable it yourself.

And considering that Instagram is owned by Meta, a company not widely recognized for its privacy-first track record - it’s worth asking how much protection truly exists beyond the message itself.

Metadata: The Data Everyone Forgets About

Earlier, I mentioned the term “Metadata.” For those of you, who don’t know what it means. Meta data can be described as “data about data.” In the context of communication, it refers to the surrounding information attached to a message rather than the message itself.

It includes who you’re talking to, when you sent the message, how often you communicate, your IP address, your device information, and other contextual details.

A common misconception involves the assumption that the content of messages is the most sensitive part of communication. But patterns can reveal just as much - sometimes more.

Imagine this - you regularly message someone named Fat Billy at 2 a.m. Almost every night. You never contact him during the day. You stop messaging him during exam weeks. You were located in a specific city when sending the messages. You delete the conversation history afterward.

When these details are mapped together, behavior becomes visible and patterns start to emerge which can be used against you. In short, Metadata allows companies or anyone with access - to model behavior and identify relationships.

And this extends beyond messaging, let’s take a look at another example.

Every time you upload a photo from your smartphone, you may be attaching something called EXIF metadata. This is hidden technical information embedded inside the image file. It can include the exact GPS coordinates where the photo was taken, the date and time, the device model, and even the time zone.

This data can be extracted quite easily using tools like ExifTool.

The John McAfee EXIF Metadata Exposure

In 2012, John McAfee who is a cybersecurity entrepreneur was in hiding. A journalist from Vice posted a photo of him online. The image still contained GPS EXIF metadata. Authorities extracted the coordinates and ultimately his location was exposed.

In short, the data attached to the picture ended up exposing him.

Communication Tools to Rethink

Let’s start with email.

Gmail and most major email providers encrypt messages in transit. That means emails are protected while traveling between servers. But that does not mean they are end-to-end encrypted. The provider can technically access message content. In corporate environments, administrators can often read employee emails as well..

SMS is even worse. Traditional text messages are not E2E encrypted. They are vulnerable to interception and are logged by carriers. They should never be used for sensitive communication.

The same caution applies to most mainstream social media messaging platforms. Instagram, Facebook Messenger, TikTok, and X (formerly Twitter) either do not provide end-to-end encryption by default - or do not provide it at all. Even where encryption exists, it is often optional and must be manually enabled.

Slack messages and Google Meet calls are not end-to-end encrypted. Zoom does offer E2EE, but it must be manually enabled through the Zoom web portal, and not all meetings use it by default.

What About WhatsApp and iMessage?

WhatsApp claims to use end-to-end encryption. Technically, message content is protected. However, it is owned by Meta, and many folks including me, remain skeptical about broader data practices. For non-sensitive communication, WhatsApp is generally acceptable. But for anything truly sensitive, stronger options exist.

iMessage is end-to-end encrypted and includes encrypted backups. However, it works only between Apple devices. You must share a phone number or email address to use it, and you may not always know whether the recipient is using iMessage or standard SMS.

Once again, for non-sensitive conversations, iMessage and WhatsApp are fine.

Also, just for your awareness - there is also an ongoing lawsuit challenging WhatsApp’s encryption claims. WhatsApp and Meta state that messages are end-to-end encrypted, meaning no one outside the chat can read them. The lawsuit alleges that Meta can access messages in near real time through internal processes.

So far, there is no definitive public evidence supporting those claims. Maybe it’s true, maybe it isn’t. If it were proven, it would represent one of the largest privacy scandals in modern tech.

Either way, transparency eventually surfaces….

Privacy-Focused Alternatives and Their Tradeoffs

If you want better privacy, you have to choose tools built with that priority in mind. Although better does not mean perfect.

Let’s look at a few widely used alternatives and where their limits still exist:

Proton Mail

Proton Mail is structured differently from traditional email providers. Its architecture is designed so that the company cannot directly access stored email content. This zero-access model strengthens user privacy but can limit certain convenience features.

Emails sent between Proton users are generally end-to-end encrypted. The service is based in Switzerland, a jurisdiction known for relatively strong privacy protections compared to many other countries. Jurisdiction matters because it determines how companies can be legally compelled to provide data.

That said, Proton Mail does not eliminate all exposure.

Email metadata such as - sender, recipient, timestamps - is embedded in the email system itself. Subject lines are not consistently end-to-end encrypted. Messages sent to Gmail or Outlook accounts lose automatic E2EE unless additional encryption steps are taken. IP addresses may still be logged unless masked through tools like Tor or a VPN. And if your device is infected with malware, no email provider can compensate for that.

Undoubtedly, proton significantly improves email privacy but keep in mind, it does NOT redefine how email works at the protocol level.

Signal

Signal is widely regarded as one of the strongest consumer messaging applications available. Unlike many platforms where encryption is optional, Signal’s end-to-end encryption is always active.

Its reputation for security extends beyond everyday users. According to public statements from CIA Director John Ratcliffe, Signal is widely used within the agency. Ratcliffe told the Senate Select Committee on Intelligence that Signal was installed on his CIA computer upon confirmation and is used by most CIA officers. That level of institutional adoption reflects trust in its security design.

Signal also allows users to communicate via usernames instead of exposing their phone numbers. Contacts can exist entirely within Signal without being uploaded to a central database in a readable form. Your phone number does not need to be visible unless you choose to share it.

Still, Signal has limits.

It requires a phone number to register, creating some degree of identity linkage. Internet service providers can see that you are connecting to Signal servers, even if they cannot read the content. Timing patterns such as - who you message and when - may still exist at a network level. And again, if a device is compromised, even the strongest encryption cannot protect the conversation.

I would recommend checking out the following post by if you want to dive more into learning stuff regarding signal:

Tate's Online Safety Substack

Think Your Messages Are Safe? If You’re Not Using the Right App, They Could Be Intercepted

Rightfully so, many Americans are more worried than ever about the interception of their personal communications and what to ensure the privacy and security of who they talk to and what they say…

Read more

a year ago · 92 likes · Tate Jarrow

Jitsi Meet

For video conferencing, Jitsi Meet presents a different model from mainstream platforms like Zoom or Google Meet. It is open-source and can be self-hosted, giving organizations full control over infrastructure if they choose to deploy it on their own servers.

For casual use, meetings can be started without creating an account, reducing identity exposure. This makes it useful for both informal and professional communication.

However, default Jitsi meetings are encrypted in transit, not always fully end-to-end encrypted in the strictest sense. Server operators can still see IP addresses and certain session metadata. Meeting duration and participant counts remain visible at the server level. As always, browser and device security play a critical role.

Accepting the Limits of the System

There is a hard truth about digital communication that no encryption tool can erase.

For a message to move across the internet, the network must know three things: where it came from, where it’s going, and when it was sent. Without that information, routing simply wouldn’t work. This is metadata and it’s unavoidable.

Keep in mind - you can’t eliminate metadata because that’s fantasy. The real objective is minimizing who can access it and how easily it can be analyzed or correlated over time.

Here are some additional steps you can take to minimize your metadata:

Hide Your IP Address

Your IP address reveals decent information about you - an approximate location, your internet service provider, and sometimes even your identity through account records.

Using a privacy-focused VPN such as Mullvad can mask your real IP from the services you connect to. But understand what’s happening - you’re placing trust on someone else, not eliminating it. Instead of your ISP seeing everything, your VPN provider will be.

There’s more depth to networks, tunnels, and traffic routing than we’re covering here. That deserves its own dedicated discussion and it will have one, stay tuned for that.

Use Apps That Minimize Metadata

Not all end-to-end encrypted services treat metadata the same way.

Some store connection logs. Some build analytics systems around usage patterns. Even the privacy-respecting tools discussed earlier aren’t flawless, nothing is.

But, here’s what you can do - Before using any service, examine how it encrypts data, what it stores, how long it stores it, and how it makes money. All of this matters more than you think.

Blind trust defeats the purpose of privacy.

Control Your Device

If your device is compromised in any way, either via malware or spyware or if someone has root access to your computer, then nothing really matters.

So, start focusing on what you install on your device, where you are downloading applications from, what you are clicking on the internet, what already exists on your device, using an antivirus like malwarebytes is also a good thing.

I would suggest checking out the following pieces to get some help on this particular matter:

Scope and Threat Model

If you are someone technical, you might wonder why this guide doesn’t dive into advanced tactics like identity compartmentalization, timing obfuscation, decentralized systems, or complex anonymity frameworks - all these techniques allow you to minimize your metadata even further.

I get it, such methods exist, they work. But they are designed for high-risk threat models.

This guide is for people trying to raise their baseline privacy, not disappear from nation-state intelligence.

A more useful starting question is this:

“Who are you trying to protect your data from?”

Because protection depends entirely on the threat model.

• Layer 1: Your Internet Service Provide (ISP)

• Layer 2: Opportunistic hackers

• Layer 3: Big Tech companies

• Layer 4: Governments

• Layer 5: Targeted surveillances

Each layer requires completely different tools, discipline, and tradeoffs to protect yourself from it.

The privacy-series I am doing focuses on layer 1 – layer 3. Government-level or targeted surveillance is a different category entirely, and defending against that requires a different level of operational rigor.

Communication Habits That Matter More Than Tools

If you made it this far, good. Most people never question how their communication systems work. You did and that is another step in the right direction.

Know this - Private communication isn’t about hiding. It’s simply about choosing the right channel for the right conversation.

Encryption, apps matter, infrastructure, all of this stuff matters, but behavior sits above all of it.

That’s why you need to start being more aware of your actions.

Be conscious of what you discuss, where you discuss it, and with whom. Not every conversation belongs on every platform nor every thought needs to live in a searchable chat history.

Separate personal and professional channels. Don’t mix identities casually. Check whether your cloud backups are encrypted. And understand this clearly - deletion does not automatically mean disappearance. Data lingers in backups, logs, and recipient devices.

Privacy is shaped by habits more than software, you can start today by working on building small, consistent habits:

• Thinking before choosing a platform

• Limiting oversharing

• Keeping your devices clean and updated

Over time, that discipline compounds.

In the next part, I will shift focus to something that is a bit more overlooked aka The Browser. It is one of the most powerful surveillance surfaces in your digital life. I’ll break down why it matters and what practical steps you can take to fix what’s leaking.

Thanks for reading.
See you next time.

The Algorithm is Winning!

I got a confession - most of my time goes into researching on topics - so, I can help you see what the system doesn’t want you to see. At this point, it’s me versus The Algorithm and the Algorithm is winning…

That’s why I need your help.

If SK NEXUS has earned a place in your inbox, consider subscribing if you haven’t already.

And if you’re active elsewhere, follow along on Bluesky or X (Twitter) or LinkedIn - those are the two places I show up most consistently outside Substack. A follow, a like, a repost - it matters to me, more than you think.

Independent research only works if the signal travels. If this work has been useful to you, help it travel.

What You Might Like

If you want to learn more, you can give the posts of different creators involving , Erich Winkler, , , on different topics a read:

ToxSec - AI and Cybersecurity

The Dead Internet Is No Longer a Theory

BLUF: The dead internet theory graduated from conspiracy to observable fact in 2024. Bots crossed the 51% threshold. Over half of new articles are synthetic. Humans detect AI text at coin-flip accura…

Read more

8 months ago · 55 likes · 61 comments · ToxSec

The Digital Watchtower's Substack

AI companions. The good, the bad, and the very very ugly.

Science fiction has predicted purpose-built cyber companions across the full media spectrum over decades. Joi in Blade Runner 2049, Iris in Companion, Samantha in Her, and even Spielberg’s 2001 Artificial Intelligence centered around Jude Law’s performance as an android sex worker named Gigolo Joe. Even children’s and teen’s media has touched on this tr…

Read more

5 months ago · 4 likes · 2 comments · The Digital Watchtower

Decoded Security

5 simple things you should do to protect your privacy before 2025 ends

The Internet is no longer secure. Was it ever? Probably not, but now we have everything online, everything digitized…

Read more

6 months ago · 15 likes · 5 comments · Erich Winkler

Heuristics vs Traps

What we gave up without noticing: how we traded safety for convenience

We like to think that history moves through shocks and explosions. Revolutions. Wars. Coups…

Read more

6 months ago · 31 likes · 13 comments · Mila Agius

Leadership in Change

Understanding Privacy in the Age of AI, a Deep Dive

Before we start: Ready to implement and build with AI? Join Premium for $1,345+ in tools and systems at $39/yr. start here | Looking for 1-on-1 coaching or a private knowledge hub build? Book a free call…

Read more

4 months ago · 36 likes · 11 comments · Joel Salinas and Mohib Ur Rehman

You must have noticed that we’re moving toward a world where surveillance is becoming the default. Every account and every device leaves a trail. Your digital life is becoming increasingly esxposed and valuable to people you never intended to trust.

So the main question is - “What can you realistically do to protect yourself?”

That’s where most people get stuck. The internet is full of advice, but almost none of it is structured. Online advice is scattered or overly technical. You’re either told to “use better passwords” or handed a wall of jargon - nothing in between.

I noticed that gap and decided to fix it.

This article is the first part of a new series focused on practical, step-by-step ways to secure your digital life, and we’re starting at the front door, aka authentication.

In this piece, I’ll break down the different types of authentication, why some are dangerous, why others are misunderstood, and how to implement them properly to strengthen both security and privacy.

If you haven’t already, I strongly recommend reading my previous piece on understanding privacy in today’s digital world. It lays the groundwork for everything that follows.

Who This Guide Is For and Who It Isn’t For

This guide is written for everyday users who want to reduce unnecessary tracking and limit how much of their data spills across the internet. It assumes you’re using normal devices and legal online services. If you’re looking for practical privacy habits you can realistically stick to - you’re in the right place. It’s also for readers who want to understand how accounts are attacked, how data gets collected, and where most people quietly lose control without realizing it.

This is not a guide for people under active state-surveillance. If your safety depends on anonymity or if you are seeking operational-security playbooks, this series is not enough.

With that out of the way, let’s start.

What Authentication Really Means

Authentication is simply the process of proving you are who you claim to be. Every time you log in, unlock a device, or confirm a prompt, authentication is happening.

And keep in mind - the password is just one possible proof you present during that process. Authentication is the broader mechanism deciding whether you are who you claim to be. The password is merely one tool in that toolbox.

This is where people often confuse two related but very different ideas: authentication and authorization.

Authentication answers the question, “Who are you?” Authorization comes after and asks, “What are you allowed to do?” Logging into an account is authentication. Being able to view files, change settings, or access admin features is authorization.

One proves identity. The other grants permission.

The reason authentication exists at all is because the Internet has no built-in way to know who you are. Unlike the physical world, there’s no face or body to recognize. Everything online is just requests flying between machines. So the internet makes a risky assumption by default: you are who you say you are - until something proves otherwise.

This is why authentication matters and why having just a password in today’s age isn’t enough.

What Happens When You Log in

At its core, a login is a shortcut for trust.

You type a username and a secret, usually a password, and the system checks whether that secret matches what it has on record. The server never stores your actual password. Instead, it stores a transformed version of it called a hash. When you log in, your password is hashed again and compared to the stored value. If they match, access is granted.

Once verified, the system creates a session. This is a temporary proof that you have already authenticated, so you do not need to re-enter your password on every page. That session is what keeps you logged in.

The fragility comes from the layers around this process. Passwords can be reused, hashes can be leaked, and sessions can be stolen. None of this means the system is broken. It means it was built for usability first, with security layered on top, not the other way around.

Some of The Main Ways Systems Verify Identity

Modern systems rely on a small set of authentication methods. None of them are perfect. Each solves one problem while introducing others:

Passwords

Passwords are the most common form of authentication and the weakest when used alone. They fail due to reuse, phishing, data breaches, and poor storage practices. Strong passwords and password managers reduce risk, but it’s important to note that they do not eliminate it.

One-Time Codes

One-time codes add a second factor, but the delivery method matters, there are two types of one-time codes:

• SMS-based

• Authenticator app-based

SMS-based codes are vulnerable to SIM-swapping attacks, where attackers take control of a phone number to intercept messages due to which they aren’t really a safe option.

On the other hand - App-based authenticators avoid some major risks and are generally more reliable, but they still depend on device security and user behavior.

Biometrics

Biometrics verify physical traits like fingerprints or face scans. They are convenient, but it’s do know that they aren’t secrets. If compromised, they cannot be changed. In practice, biometrics usually unlock a device that still relies on another authentication method underneath. Banking applications are a great example of this.

Anyway, the key takeaway is tradeoffs. Each method improves usability or security in one area while weakening it in another - so yeah, nothing is perfect just by itself.

How Authentication Actually Breaks in the Real World

No system is unbreakable. Authentication looks solid on paper, but real-world usage introduces human error, reused credentials, and predictable behavior. Due to this dozens of accounts get hijacked nearly everyday. And it’s not that the systems aren’t secure enough, this happens because attackers target users.

Most compromises follow a handful of well-understood paths, repeated at massive scale. Let’s look at the most common ones:

Phishing

Phishing is a method in which attackers trick users into entering their credentials on fake login pages that look identical to real ones. Once the credentials are captured, the attacker logs in as the user.

The damage multiplies because of password reuse. Many users reuse the same password across multiple services. According to Cloudflare’s analysis, nearly 41% percent of login attempts involve credentials that were already exposed in previous breaches. That means one successful phish can unlock multiple accounts.

This is why phishing remains effective even against secure systems.

Phishing, along with several other account-takeover methods, falls under social engineering. If you want a deeper breakdown of how these attacks manipulate behavior rather than technology, you can read the full explainer here:

Multi-Factor (MFA) Fatigue

Multi-factor authentication is meant to stop attackers even if passwords are stolen. MFA fatigue attacks flips the entire game.

Also known as MFA bombing or MFA spamming, this attack involves repeatedly sending authentication prompts to a victim’s phone, email, or authenticator app. The attacker’s intention is to somehow exhaust the person on the other end. Eventually, out of annoyance, confusion, or urgency, the victim approves one of the requests.

That single approval is all the attacker needs.

In practice, the attack usually unfolds in three steps.

• First, the attacker obtains valid login credentials, often through phishing

• Next, they attempt to log in repeatedly, triggering a flood of MFA push notifications.

• Finally, the victim becomes frustrated and taps “Approve,” unintentionally authenticating the attacker.

Yeah, this is it, no fancy hacker-guy exploit goes on behind the scenes.

Account Recovery Abuse

Account recovery attacks come into play when authentication is blocking them or if they don’t have the credentials.

Instead of breaking into an account, attackers abuse password-reset and recovery flows - the very mechanisms designed to help legitimate users regain access.

At the core, these attacks rely on impersonation and social engineering. The goal is to ultimately convince the system you are who you claim to be, and the protections meant to stop unauthorized access collapse.

This plays out in many real breaches. For example, a major ransomware attack against Clorox was traced back to attackers impersonating employees and talking a corporate help desk into handing over credentials.

There are countless variants of account recover attacks but some common flavors of recovery abuse include:

• Security question guessing

• Support desk social engineering

• Email takeover (once an attacker controls your email, every linked account becomes vulnerable via password reset links.)

Why Understanding All this was Necessary…

You might be wondering why I am not getting to the point.

But what you need to understand is, knowing all this - is a part of the process. I am laying out these attack patterns first because you must guard your own mind first.

By now you must have realized that most compromises don’t defeat security technology. They exploit trust, urgency, and poorly designed recovery paths. And understanding all these different attacks is the first step toward securing your digital life and protecting your privacy in a system that rarely does it by default.

Designing Authentication That Works in Real Life

Up to this point, we’ve focused on how authentication fails in the real world - it was to show you a simple truth - most account compromises don’t happen because the technology is weak, rather it’s the fault of the setup itself. (though other factors are play a key part in it as well)

But, this is where we shift from problems to practice.

A reasonable authentication setup doesn’t aim for perfection. It aims to raise the cost of attack while keeping your daily life usable. The goal is to block common failure paths without turning security into a full-time job.

There’s a key idea worth repeating from the last piece - being secure is necessary to protect privacy, but it isn’t sufficient on its own. You can have strong authentication and still leak data through the platforms you trust. That’s why this section focuses on pairing good security hygiene with more privacy-respecting choices.

In the sections ahead, I’ll walk you through how to build a setup that resists the most common attacks, limits unnecessary exposure, and scales with your digital life.

With that being said, let’s start.

Password Manager Setup

Weak passwords and password reuse are two of the most common reasons accounts get compromised. A password manager addresses both by generating strong, unique passwords and storing them securely.

I personally use Bitwarden and have had a reliable experience with it. It’s open-source, works across devices, and gives you control over your vault. That said, the important part is using a password manager you trust and will actually use.

Following are the steps you can follow:

1. Go to bitwarden.com and sign up.

2. Create a strong master password.

   • Long and unique.

   • Include letters, numbers, and symbols.

   • Never reuse it anywhere else.

3. Install the browser extension or visit the site to login and access your vault.

4. Let the manager generate passwords for new accounts.

5. Over time, update old logins with reused or weak passwords.

Your master password is the single point of access to your vault. Treat it seriously. If it’s weak or reused, the rest of your setup doesn’t matter.

App-Based Authentication

Strong passwords are the first layer. Multi-factor authentication (MFA) is the second.

MFA means that even if someone gets your password, they still can’t log in without proving it’s you - usually through a time-based code generated on your device.

There are two common types: SMS-based and app-based authentication.
Avoid SMS-based MFA. It’s vulnerable to SIM-swap attacks and number hijacking. App-based authenticators are more reliable and harder to abuse.

I personally use Ente Auth and recommend it. It’s privacy-focused, encrypted, and works across devices.

Following are the steps you can follow:

1. Install Ente Auth on your phone.

2. Create an account and secure it with a strong password.

3. Back up your recovery keys somewhere safe.

4. Go to an account you want to protect.

5. Open Security or 2FA settings.

6. Scan the QR code using Ente Auth.

7. Save the backup codes provided by the service.

From now on, logging in will require both your password and a rotating code from your authenticator.

If you ever receive MFA prompts you didn’t initiate, don’t approve them. That’s often a sign of an MFA-fatigue attempt. Change your password immediately and review recent login activity.

Biometric Authentication

Biometric authentication uses something you are instead of something you remember. On phones, this usually means fingerprint or face recognition.

First, enable it on your device:

• Android: Settings → Security & Privacy → Biometrics → Fingerprint / Face

• iPhone: Settings → Face ID / Touch ID → Turn it on and register

This lets apps ask your phone: “Is this really the owner?”

It’s important to note that not every app supports biometrics. Although the support has improved over the years, it is still inconsistent. So don’t overthink it and do this instead:

• Check your banking apps

• Check email, password managers, and social apps

• Anything that holds sensitive data

If biometric login is available, turn it on.

From an average user’s POV, biometrics adds friction against casual misuse and shoulder-surfing. Just keep in mind that biometrics should be used to tighten access, not to replace passwords or authenticators

Recovery Planning: What to do When Things Go South

One of the most overlooked parts of authentication is having a recovery plan - knowing what to do before things go wrong.

Most people focus on logging in securely, but very few think about what happens if they lose access or trigger suspicious activity. Recovery planning is about making sure that when something breaks, you can regain control without weakening your security or privacy.

There are a few simple steps that make a big difference.

Backup Access

Any authentication system without backups is fragile.

When you enable multi-factor authentication, services usually provide backup or recovery codes. These exist for one reason: if you lose access to your authenticator app or device, these codes are your way back in.

• Save your 2FA backup codes when you enable MFA

• Store them securely, ideally in your password manager or offline

• Don’t rely on “I’ll deal with it later” because convenience eats away at both security and privacy

If you lose your authenticator and don’t have backups, recovery becomes painful. In many cases, it means being permanently locked out of your own account.

A few minutes spent securing backups can save you from a lot of pain.

Setting and Securing the Recovery Email

Your recovery email is often more powerful than your password. If someone controls it, they can reset passwords across your accounts without any fancy hacking.

Here are some basic rules to follow:

• Use a separate email account for recovery

• Protect it with a strong password + app-based MFA

• Avoid logging into it on shared or any untrusted devices

Treat this inbox like a master key. Lose control of it, and everything else follows. And If you’re unsure which email providers to use, don’t worry - I’ll cover that in the upcoming privacy sections.

Periodically Review Recovery Settings

Recovery setups decay over time. Your old emails, unused phone numbers, exposed backup codes - all of these can be used against you if you don’t take care of it.

Build a habit to do some sort of check-in every month that follows this outline:

• Checking recovery emails and phone numbers

• Removing outdated or unused options

• Regenerating backup codes if they’ve been exposed

This part is really important, so don’t ignore it just because it seems not so important.

Password Rotation After Suspicious Activity

You must have heard that quote - ‘Hesitation is defeat’

Well, hesitation is also what attackers rely on.

There are a couple of red-flags that should make you instantly take action. Such red flags can include any kind of unexpected login alerts, MFA prompts you didn’t initiate or even password reset emails you didn’t request.

And if you do end up spotting something like that, then you need to act immediately and do the following:

• Change the password

• Rotate it everywhere if it was reused

• Kill active sessions and log everyone out

Waiting to ‘see what happens’ is how small incidents turn into full account takeovers.

And here’s another tip - Visit Have I Been Pwned. Drop in the email you’re worried about and see if it’s tied to known breaches. It’s not perfect, but it often gives you an early warning and that’s enough to act before damage spreads.

Why Authentication Alone Isn’t Enough

If you’ve made it this far, that’s already a win.

You’ve taken the first real step toward securing your digital life instead of reacting after something breaks.

But authentication is just the beginning.

Strong passwords, MFA, and recovery planning protect the door. What happens after you’re inside is a different problem entirely.

In Part 2, we’ll move beyond login screens and talk about communication - how emails, messages, metadata, and everyday digital interactions expose you even after authentication succeeds.

I’ll break down the basics, show where the real leaks happen, and walk you through practical ways to tighten things up and reclaim some privacy.

Thanks for reading.
See you next time.

The Algorithm is Winning!

I got a confession - most of my time goes into researching on topics - so, I can help you see what the system doesn’t want you to see. At this point, it’s me versus The Algorithm and the Algorithm is winning…

That’s why I need your help.

If SK NEXUS has earned a place in your inbox, consider subscribing if you haven’t already.

And if you’re active elsewhere, follow along on Bluesky or LinkedIn or X - those are the two places I show up most consistently outside Substack. A follow, a like, a repost - it matters to me, more than you think.

Independent research only works if the signal travels. If this work has been useful to you, help it travel.

What You Might Like