Hey there 👋
Welcome to another issue of Arc - the writing sandbox.
It is clear that AI will become the default feature on many products, and in return part of many workflows. If you still choose to do things without making use of it, upto you, and more power to you - but for many, especially in fast paced environments - we don’t really have a choice now, do we.
I still like working with hand, typing with mind, thinking by inspiration, and I still find peace in just day dreaming.
I like taking my glasses off and just looking into nothingness with my -5.00 bad eyes.
I like to defocus, and enjoy the blur, and just gaze without conscious choice every now and then.
But
When its time to get executin - AI is now part of many workflows of mine.
In today’s article I want to go over how you can too make healthy use of it - without joining the dark side completely.
I’m a big believer in first principles thinking.
Strip away assumptions. Break a problem down to its fundamentals. Rebuild from the ground up.
It’s how I approach most things - projects, decisions, systems, even my own habits. So it was only natural that when people kept asking me, “How do you actually use AI in your workflow?” and I found myself fumbling the answer, I went back to the drawing board.
Not to look for trendy tools or productivity hacks.
But to ask a simpler question:
What is AI actually good at?
Not “what can AI do,” but “what should I use it for”, especially if I’m trying to do thoughtful work across writing, design, research, planning, or product building?
That’s when a clear pattern emerged.
The highest-return use of AI wasn’t replacing entire workflows.
It was handling what I now call “The First Step”.
What Is “The First Step”?
Think of it as a practical fork of first principles thinking.
First principles thinking asks: What are the core, foundational truths of a problem?
The First Step asks: What’s the most friction-heavy, inertia-ridden starting point of a task - and how can AI help me get past it?
It’s about using AI to eliminate the cold start. The blank page. The empty Figma board. The awkward kickoff doc. That uncomfortable “where do I even begin?” feeling.
Once you see it this way, the use cases start pouring in:
Writing an article? Let AI sketch a structure and potential themes.
Designing a UI? Prompt AI to show layout inspiration and UX patterns.
Planning user research? Ask AI to surface domain insights and sample questions.
Prepping for a client meeting? Let AI simulate objections or generate a prep doc.
Starting a product strategy? Use AI to break down the competitive landscape.
The First Step isn’t about handing over the whole task. It’s about giving yourself a head start.
Why The First Step Works So Well
We’re wired for iteration, not genesis.
Most people - even talented ones - are far better at critiquing, editing, and improving something than they are at conjuring that “something” from scratch.
That’s not a flaw. That’s just how cognition works. It’s easier to react than initiate. Easier to shape than summon.
AI is uniquely powerful here because it provides that reactable surface. It gives you something - anything - to start with. And in a world drowning in ambiguity, having a rough version is infinitely better than none.
Before AI, your first step often looked like:
45 minutes staring at an empty Google Doc
Opening 12 tabs and getting distracted by 9 of them
Rewriting your Notion task title 5 times
Now?
You can start with a draft. A stub. A prototype. A skeleton. That one little jumpstart changes everything.
What AI Can’t Do (And Why That’s Important)
Here’s where a lot of the AI hype falls apart: assuming the tool can do the whole job.
It can’t. And it shouldn’t.
The First Step framework only works if you bring something to the table.
That something is your:
Taste - to know what’s good and what’s garbage
Context - to spot what’s relevant and what’s noise
Experience - to refine rough ideas into real ones
Intuition - to know what’s missing even when it isn’t said
AI doesn’t replace these. It amplifies them.
If you don’t know where to go, AI can’t get you there. But if you know the general direction and just need a push to start walking, it’s your best shortcut.
That’s why this framework isn’t about automation. It’s about momentum.
Real Examples from My Workflow
Here’s how I apply The First Step approach in real scenarios.
Writing
Before: Blank doc, scattered thoughts, hard to prioritize.
Now: I prompt AI with a one-liner - “Write me a 5-section outline for an article about why newsletters fail in Pakistan.”
I rarely use the exact structure, but it gives me a starting point to disagree with or shape.
UI/UX
Before: I’d scroll Dribbble or Behance for hours.
Now: I ask AI to generate mobile screen layouts for specific flows - e.g., onboarding screens for a travel app with dark mode.
I get rough sketches or visual prompts that guide my next move in Figma.
Research
Before: I’d have 20 open tabs and no coherent summary.
Now: I use AI to synthesize Reddit threads, academic papers, or industry articles into pros/cons, definitions, or insight clusters.
and the list goes on.
See, I work on A LOT of different kinds of tasks. For me, any given day I will be running through ‘The First Step’ at least 5-6 times.
I work as a PM for companies, I consult dev companies, I run a community, I write here, I help others write over there, I have to manage comms, I have to create documentation, any given day, I would be typing away at my computer for at least 2-3,000 words (that’s just me, no AI).
What You Should Try Next
Next time you’re about to start something that feels large or hard or fuzzy, stop and ask:
“What’s the first step here that AI could help me not do alone?”
If you’ve got the domain knowledge, AI can be your sketchpad.
If you’ve got the intent, AI can give you your first artifact.
But if you’re lost entirely, AI won’t save you. You need to bring direction. It brings motion.
This framing helps you cut through the noise of endless AI apps and focus on high-leverage use: reduce friction, accelerate iteration.
The blank page doesn’t have to be your enemy anymore.
Why This Isn’t a Hack - It’s a Mental Model
The First Step isn’t a prompt trick. It’s a way of thinking.
Most AI advice either falls into cheerleading (“AI will change the world!”) or fear-mongering (“You’ll lose your job!”). Neither helps you actually use the tools better.
But this does:
Map the task. Break it into logical stages.
Identify the hardest starting point.
Deploy AI only for that first friction point.
Iterate with human judgment from there.
If you’re a mid-career pro trying to stay sharp, or a tech-curious student figuring things out - this mindset lets you build with AI, not just consume AI tools.
AI won’t do the hard work for you.
But it can make starting less hard.
The rest? That’s still on you.
With or without my help – I wish you the best.
The Wandering Pro is a quiet corner of the internet where freelancers, tech workers, and first-time builders gather to make steady progress - one challenge, one win, one project at a time.
Join The Wandering Pro; find your rhythm, share progress, and grow with a community backed by decades of real-world building experience.