I can't write code. Five months and more than five thousand changes later, I still can't. What I could do was sit down almost every night and describe, as precisely as I knew how, a world I wanted to exist. An AI tool called Claude Code did the rest. That world is called BeMatrx now: ninety cities, elections, a working economy, twelve languages. All in, it cost me about $1,100 and most of my evenings. This is the honest version of how it happened, including the night everything broke.
Who's telling this story
My name is Sezgin Çetin. I live in Antalya, Türkiye. Until last year I had never opened a terminal, never sat through a computer science class, and honestly believed a "repository" had something to do with banks. I'm not telling you this to be charming. It's the starting line, and the starting line is the whole point: everything below happened without me learning to program.
There was one thing I'd always been good at, though I never thought of it as a skill. I could describe things until they were unmistakable. Most of adult working life is that anyway: you explain what finished looks like, the work comes back, you name the gap, it goes out again. Nobody ever told me that's also most of what building software is. The typing turned out to be the only part I was missing.
The itch
The idea arrived as a question I couldn't put down: what would an app feel like if the time you put into it left something behind? A career that actually progresses. A friendship that deepens because you both keep showing up. A city on the other side of the world that slowly becomes yours. I wanted to be able to close a screen at night and have something to show for the evening. It really wasn't more complicated than that.
I had no team, no budget a startup would recognize, and no way to build any of it myself. What I had was a suspiciously clear picture of what I wanted, and a growing hunch that the picture was the hard part.
Meeting Claude Code
A friend mentioned Claude Code to me almost as trivia. Anthropic's coding agent, he said. People talk to it and it builds things. I assumed exaggeration, because most sentences with "AI" in them are. Then one night I opened it, described a welcome screen the way I'd describe it to a designer, and watched it appear. Working. Then it explained what it had done in plain sentences, like a contractor walking me through a room he'd just finished.
Two things about it still surprise me, and I've had five months to get used to them. The first is that it never once made me feel stupid. I asked what a database was. I asked where the data physically lives. I asked, at two in the morning, whether deleting a file was dangerous. Every time, I got a straight answer instead of a sigh. The second thing mattered more: it argues back. When I asked for something that would have quietly broken the world three weeks later, it said so, and explained why in words I could actually follow. I have paid consultants a great deal of money for a great deal less honesty.
What a night actually looked like
The routine was never "learn to code, then build." It was describe, look, correct, repeat. Kitchen table, tea going cold, my phone propped against a bowl so I could test on the real thing. Here's an exchange from early on, close to word for word:
Me: When someone opens the city map, it should feel like arriving somewhere, not like a page loading. Give it a breath before anything appears.
Claude Code: builds it, then explains in plain language what changed and why.
Me: Closer. But on a slow connection that pause will read as frozen. Show a small sign of life right away, and delay only the reveal.
Months later, the conversations had grown up with the world. This one is roughly how elections came to exist:
Me: I want a city to be able to elect a mayor. Real votes, a real term of office, a real title you keep afterwards.
Claude Code: proposes how it could work, closes the obvious cheat of voting twice, and then asks three questions I had no answer to: what happens to the candidates who lose, what happens if nobody runs, and how long a term should last.
Read that second one again. The tool was asking better product questions than I was, about my own product. It kept happening until I stopped being surprised and started counting on it: whatever I brought to the table at nine, the conversation had sharpened by ten.
And no, I never read the code. Not once, not really. I read results, argued with behavior, and moved on. People find that the hardest part of this story to believe. It's also the truest sentence in it.
The night everything broke
There's a night I still think about. An update that had run perfectly on my phone an hour earlier refused to open on a friend's. Just a white screen where a city should have been. I didn't know what a stack trace was. I didn't even know what to call the problem. So I did the only thing I knew how to do, which was describe the symptom as if everything depended on it. What I tapped. What I expected. What I saw instead. Which phone, at what time, and what had changed that day.
It was enough. Not because I got lucky, but because describing symptoms precisely is a real skill, and it transfers from ordinary working life straight into building software. You don't need to know why the oven died to say exactly what happens when you turn the dial. We found it, we fixed it, and I walked away with the rule that shaped the rest of the project: I don't have to understand the machinery. I have to be relentless about what's wrong and what "fixed" looks like.
The first time it felt real
The moment I actually believed in this wasn't a launch or a milestone. It was smaller. Someone I had never met opened the app, chose a city they had never been to, and sent a first message to a stranger living there. I watched it happen live, and felt the ground shift a little. This wasn't my app anymore. It was a place, and somebody had just moved in.
What it cost
People always want the number, so here it is: about $1,100 across the five months. Roughly half of that went to Claude Code itself, and I'll say plainly that it stopped feeling expensive in the first week, right around the time I compared it to what a single week of a contract development team would have cost me. The rest was cloud hosting and a database that grew as the world did, the Apple and Google developer accounts every app pays for whether it's built by forty engineers or one stubborn man at a kitchen table, and a small pile of domain, email, and odds and ends that nobody remembers to count when they ask what building an app "really" costs.
Not free. Not nothing. But a year earlier I would have told you, with complete confidence, that the thing I wanted to build required a team I could never afford.
If we were having coffee
You'd probably ask what I actually learned, so let me answer the way I would across a table instead of like a slide deck.
You don't need to read code to reason about a product. You need to know exactly what you want and be able to say why, and that skill was never locked behind a computer science degree. Precision beats vocabulary, every single time; I never once needed the correct technical term for a bug, only the correct description of what happened. Momentum beats perfection, because the small, slightly wrong version that ships teaches you more than the flawless version that doesn't exist. And the hard part was never the tool. It was saying no. Every feature I refused made the ones I kept feel more like a place and less like a menu of buttons.
Somewhere around month four, I noticed I had stopped saying "the thing I'm trying" and started saying "the world I'm building." Five months, it turns out, is exactly long enough to stop calling something an experiment.
One more, for a specific audience. If anyone at Anthropic ever reads this: somewhere in your logs there is a man asking, at two in the morning, whether deleting a file is dangerous. Thank you for building a tool that answered him kindly, and for making "I can't code" the beginning of a sentence instead of the end of one.
If you're new here
BeMatrx is a living world, not a feed. There is no algorithm deciding what you see, and that was a decision, not an oversight. You pick a city, a career, a version of yourself, and you live the life you choose, with real people. The doors are opening gradually, and this blog is where I'll keep telling the truth about what building it is like.
