Building in the Age of Machines

There’s a question that keeps coming back to me, usually late at night, usually when I’ve spent the day working alongside AI tools that can do in seconds what used to take me hours.

The question is this: When the machines can build, what does it mean to be a builder?

It’s not a new question. Every generation of tooling provokes it. The printing press didn’t kill writers. The camera didn’t kill painters. The synthesiser didn’t kill musicians. But it did change what it meant to write, to paint, to play. It changed the point of the thing.

The craft wasn’t the point

Here’s what I think most people get wrong: they assume the value of building was always in the mechanics of it. The typing, the debugging, the syntax, the boilerplate. The craft, in the narrowest sense.

But craft was never the point. Craft was the cost of entry. It was the thing you had to get through to get to the thing that actually mattered — the thinking. The decisions. The taste. The judgment about what to build and why and for whom.

If AI lowers the cost of entry, that doesn’t diminish the builder. It reveals them. It strips away the false signal of mechanical skill and exposes the real signal: Do you have something worth building?

The new literacy

I think we’re heading toward a world where the ability to direct AI — to think clearly about problems, to articulate intent, to evaluate output — becomes a form of literacy as fundamental as reading and writing.

Not prompt engineering. That’s a transitional skill, like knowing how to hand-crank a car engine. I mean something deeper: the ability to think in systems, to hold complexity, to know what good looks like even when you didn’t personally assemble every piece of it.

What remains

The machines will get better. They’ll write more of the code, generate more of the designs, handle more of the implementation. That’s not a threat. It’s a liberation — if you have something to say.

The builders who thrive won’t be the ones who can type the fastest or memorise the most APIs. They’ll be the ones who know why they’re building, who carry a vision that no amount of autocomplete can generate.

The tools change. The point doesn’t.