Built to build

I’ve been building things for as long as I can remember. Lego as a kid. Warhammer worlds as a teenager – though I was terrible at painting the miniatures. Fanzines, a coffee table photo book, my own black and white darkroom. My first website in the late 90s. Design systems for newspapers and magazines twenty years ago. Companies. Digital products. Last year, a summer house. This year, AI prototypes.

The materials change. The instinct doesn’t.

When I pick something up, I go deep. Not just enough to get by – enough to teach it. I learn how it works, why it works, and when it doesn’t. This has been true for every medium I’ve touched, from analogue photography to component-based design systems. I don’t collect skills for the sake of range. I collect them because building demands it. Every new material teaches you something the previous one couldn’t.

I build because I want to make things that last. Things that work and look like someone cared. A photo book you keep on your shelf for twenty years. A design system that holds up across teams and products. A summer house that stands solid. I also build to belong – the best projects have always pulled people in, created friendships, sparked something shared. Building is rarely a solo act, even when it starts with one person’s obsession.

But there’s always been a gap. I think in products and experiences. I see the whole picture – the flow, the structure, the logic. I’ve spent years as a product owner, working with designers and developers to turn ideas into something real. The craft of translating that vision, though, required handing pieces of myself to others. The designer made it visible. The developer made it work. I sat between them, holding context together across meetings and handovers and Slack threads.

I’m not complaining about collaboration – that’s where the best work happens. But there was always a distance between what I could see in my head and what I could build with my own hands. I could sketch. I could write specs. But the actual building? That belonged to someone else.

AI changed that.

Not in the way the hype suggests – I’m not suddenly a full-stack developer. But for the first time, I can go from idea to working prototype without waiting. I can think through a product flow by building it, not by describing it in a document and hoping the description survives three handovers. The designer in me and the product thinker in me finally have a building partner. AI doesn’t replace the team. It extends me. It makes me more whole.

That word keeps coming back: whole. For years I’ve had strong instincts about how products should work, shaped by decades of building in different materials. But I couldn’t act on those instincts when building digital products. Now I can get close enough to test my thinking, break it, and learn from it in minutes rather than days.

This matters because curiosity without building is just speculation. I’ve never been interested in learning for the sake of learning. I learn so I can make something. AI makes that possible again in a domain where I’d been limited to thinking and talking and writing documents about what someone else should build.

I’m not interested in the hype. Generating funny images or viral videos isn’t the point. The point is whether this actually makes the work better. After three weeks of building AI prototypes, I believe it does – not because the output is polished, but because the feedback loop between thinking and making nearly disappears. That’s where the real value lives.

I’ve been building things my whole life. The tools keep changing. The drive doesn’t. Right now, I’m building with AI, and it connects parts of my experience that have been waiting to come together.