In November 2025 I sat down to build something myself, end to end, for the first time in my life. Not brief an engineer. Not manage a team and hope they understood the idea. Build it.

The excitement lasted about three hours.

Research, market validation, the shape of the thing, all of that came together in an afternoon. Then I opened the no-code platforms everyone on my feed was raving about and tried to turn the idea into a working product. The front end came together fast. The moment I needed data to move from a form into a database and back out again, safely, everything fell apart. I couldn't finish a single thing.

Meanwhile my feed was full of people claiming they'd shipped whole products in a weekend. I was sitting in Ibiza unable to connect a form to a database.

The short answer is yes

Today I build without writing a line of code by hand. This website, BigMoneyIdeas, MyWebsiteSucks and The First 100 are all live, and I built every one of them myself with AI doing the typing. So yes, a non-technical founder can build a real product with AI. I'm the proof, and I was about as non-technical as founders get.

But the honest answer is longer, and it's the longer answer that decides whether you finish.

The hard part was never the code

When I stepped away from those first attempts, I spent a few weeks trying to understand why they'd failed. Not which tool to use next. Why.

The answer was humbling. The tools weren't the problem. I didn't understand how a product works underneath. Where data lives. How it moves. What needs protecting, from whom. Which part of the system decides what, and what happens when two things try to happen at once.

AI can write the code for all of that. What it can't do is make the decisions for you, because it doesn't know your business, your customers or your risks. If you can't describe what the system should do, clearly and in order, the AI will happily build something that looks right and quietly isn't.

You don't need to learn to code. You need to learn to think in systems. That's a different skill, and it's much faster to learn than programming, but nobody teaches it in the demos.

You don't need to learn to code. You need to learn to think in systems.

The $2,000 lesson

The second thing nobody warns you about is cost.

Early on I spent $2,000 in a single week on API calls. Not because I built anything big. Because my prompt was bloated, and every request was dragging thousands of unnecessary words along with it. Nobody told me. The bill did.

Every non-technical founder building with AI will hit a version of this. A runaway loop. A model far bigger than the job needs. A database query that works beautifully with ten users and falls over at a thousand. None of these are hard to fix once you know they exist. All of them are expensive to discover alone.

Why a weekend isn't enough

The weekend-build stories aren't lies, exactly. They're demos. A demo proves the idea can exist. A product has to survive real people using it wrong, at odd hours, with their own data.

The distance between the two is where most first-time builders give up. It's not one big wall. It's a hundred small ones, each of which takes an hour to get past if you know what you're looking at and a week if you don't.

That's also why no single session, workshop or call can build a product for you. Ninety minutes with someone who knows the way is enough to set direction, untangle a decision and get you unstuck. The building itself happens in the hours between: the homework, the follow-up, the evening you get one feature working and the morning you find out why it broke.

I built AI Co-Working around exactly that rhythm. Live sessions set the direction. Between them we keep building together, you work through the homework, and you're never more than a WhatsApp message away from getting unstuck.

What you end up owning

The reason to build it yourself isn't to save money, although you often will. It's what you're left holding at the end.

You own the product, every line of it, not a subscription to someone else's platform. More importantly, you own the understanding. The next thing you build takes a fraction of the time, because the hundred small walls are behind you now.

My first attempt produced nothing. Every build after it came faster than the one before, and the difference was never a better tool.

It was knowing what I was asking for.