Collaboration Through Code
For a long time, code was a wall. You either spoke the language or you waited. And if you waited, you translated — your ideas into tickets, tickets into estimates, estimates into sprints, sprints into something that didn't quite match what you had in your head. The wall didn't just slow things down. It changed what got built.
I spent a lot of years on the design side of that wall. I understood the problem better than most people in the room, but I couldn't touch the solution. I could describe it, sketch it, argue for it — but then I had to hand it off and hope the translation held.
What's changed isn't that the wall is gone. It's that it's become optional.
I was in a room recently with a group of designers. Nobody would call themselves a developer. We opened VS Code, opened Claude Code, and started building. In an hour, people had made a Pokédex, a personal data management tool, a native mobile prototype. Real things. Things that worked. The absorption in the room was total — because they were finally touching the thing, not describing it.
That's what I mean by collaboration through code. Not that everyone should learn to code. Not that AI replaces engineers. But that code is becoming a shared medium — something you can work in together, the way musicians work in sound or architects work in space. And when more perspectives can touch the material, the material gets better.
The best products I've seen built in the last two years weren't built faster because the engineers got faster. They were built better because the distance between the person who understood the problem and the person shaping the solution collapsed. The domain expert was in the code. The designer was running the prototype. The product thinker was changing the logic directly, not filing a ticket.
There's a word for when a medium becomes accessible enough that non-specialists start working in it: democratization. But that word makes it sound like a policy. What actually happens is more interesting. When more kinds of people work in a medium, the medium itself changes. Jazz happened when brass instruments became cheap. Punk happened when three-chord progressions became teachable. The medium doesn't just reach more people — it goes somewhere new.
I think that's what's happening with code right now. And I think the people who are going to shape where it goes aren't just the engineers. They're the designers who finally get to build what they imagine. The operators who can finally fix the thing that's been annoying them for three years. The domain experts who can finally make the tool that only they knew was missing.
The wall is coming down. The question is what we build on the other side.