finding your moat

finding your moat

We’ve all been using AI heavily for coding, and it’s been amazing. I hope it’s been the same for you.

The pace has never been this fast. Hundreds of products hitting the market every single day. I’ve been thinking about it a lot. Me (a software engineer, an AI engineer): I’m offshoring most of my coding to AI now, and it’s left me with this hollow space in time. Yeah, I still do code review, I still do architecture. But a small feature? AI does that fast, and does it well. The code still needs a serious review, sure. But here’s the thing I keep landing on: in the next two or three years, AI is going to be maybe 100x better at coding than it is today.

So I keep asking myself: what’s my moat here? Was coding ever really my moat?

It was. But coding is getting commoditized now, and it’s hard to build any leverage on top of a commodity.

And I think that’s the thing we need to sit with. Coding was never your only moat. Your real moat was your system thinking, your approach, your taste, your ideas.

Because AI isn’t just helping developers, it’s turning everyone into a builder. The generalists are the ones shining now. Look at it from a non-dev’s point of view: who’s actually shipping fastest? Often it’s the solo founders, the ones who aren’t even that technical. The technical barrier is gone. Anyone can build. Product managers are happier than they’ve ever been. This is their moment to shine. I call this the building war phase, and it’s forcing all of us to become generalists. We’re all slowly turning into PMs of our own code.

But here’s the catch. If everyone can build, and everyone is shipping, what actually separates the winners from the losers?

Let me back up for a second. Yes, AI lets everyone build faster. But that comes with a cost. Look around and you’ll see people shipping random things, random features, just because they can. It’s creating clutter. Feature clutter. It’s like handing a kid a flying carpet who has no idea how to fly it. And I think that’s just where we are right now: we got handed this superpower overnight, and we haven’t figured out how to actually use it. So for now, it mostly makes chaos.

Which brings me back to the real question: what separates winners from losers?

Taste and distribution.

Taste is knowing what’s actually needed. Knowing which feature will genuinely make a dent in someone’s life, and which one is just noise. AI commoditizes the instant: writing code, writing copy, anything you can pull out of a prompt. What it can’t commoditize is the compounded: the things you build up slowly, over time, that no prompt can hand you. Taste is exactly that. It’s what tells you what to build, and just as importantly, what to kill.

Then there’s distribution. Everyone’s building, but has your product actually reached anyone? Has it made a dent in a single real person’s life? You can rent attention with ads and campaigns, but the moment you stop paying, it’s gone. A real audience is earned slowly, over time. AI can’t conjure that audience for you any more than it can conjure your judgment. Distribution is what decides whether the thing you built lives or dies.

So in this building war, the carpet is free. Everyone gets one. But what you do with it is still entirely on you.

There’s a line I keep coming back to: “what is this earth without art? Just a solid rock.” That’s what taste is now. It’s your art. And your art is the one thing nobody can copy. AI least of all.