You are the bottleneck.

For two decades the wall between an idea and a working tool was code. Around mid-2026, for anyone who can describe what they want in plain English, that wall is gone.

For two decades the wall between an idea and a working tool was the same brick: code. Around the middle of 2026, for anyone who can describe what they want in plain English, that wall quietly came down.

For the longest time, anyone who wanted to build something in the digital space hit the same roadblock — execution. Building software was hard, and it kept getting harder. New languages, new frameworks, faster cycles. Keeping up with even one of them was a full-time job.

Anyone like me, who either never started coding or lost interest somewhere around Visual Basic, got left behind. The hurdle to enter the game looked so big that it simply wasn't an option. That stopped me — and presumably a lot of other people — from ever executing.

The ideas were never the problem. The wall was always the code.

Two decades of unbuilt ideas

For twenty years I've had no shortage of ideas: tools to build, software to build, ways to fix the tools I was already using. Most of them were probably useless or unworkable. But say two percent weren't. I still couldn't touch them, because I couldn't build them. I'm not a coder.

Then the no-code wave arrived. Website builders first — Wix, Squarespace — made it genuinely easy to put something on the web without writing a line. But that was mostly marketing sites sitting on a CMS. Not functioning apps. The wall had a door now, but only into one room.

The wall comes down

The real shift started around 2025, when it got out that developers weren't just using AI to check their code — they were using it to write it. At first that was still out of reach for the ordinary person. Not anymore.

Now, halfway through 2026, the last hurdle is gone. For roughly a hundred francs a month and the ability to express yourself in English, you can build almost any tool you want. Claude Code, OpenAI's Codex, and the rest have made it possible to build things without knowing how to build things — arguably a first in the history of computing.

What's left

For someone who's always had ideas but never been able to act on them, this is a paradigm shift. I can now execute on basically any idea I have.

Which doesn't mean I should — but that's a thought for next time. Can build and should build are two different questions, and the gap between them is where most of the junk gets made. The wall coming down is a gift; it's also a licence to make a mess.

So there are still limits — technical, financial, and the ones worth keeping. No, I can't build every tool in my head, and I shouldn't try. But when an idea has stuck around for weeks or months, and I still think it could be useful, or even worth selling, I can finally build it. Whether it earns back the time I put in is a separate question, and a real one.

But for the first time, when it comes to building digital tools, the only bottleneck left is me. My imagination.