Where value goes when building is free

16 Jul 2026

AI is driving the cost of writing software toward zero. It's tempting to read that as "developers are finished." I read it the way an economist would: when the cost of one input collapses, value doesn't evaporate. It moves to whatever just became the new bottleneck. The interesting question isn't whether building gets cheap. It's what gets scarce when it does.

When a cost goes to zero, value moves to the complements

This is an old pattern. When one part of a system gets cheap, demand and value shift to the parts that are now the constraint. When building is abundant, the scarce things become: judgment — of everything that can be built, what actually should be, and for whom; verification — once generating an answer is free, knowing whether it's correct is the hard part; and the last mile — turning a capability into something that runs and is trusted in a specific, messy reality. "AI flattens the price of code" isn't the end of the story. It's value moving house: out of output, into judgment and verification.

The thing AI structurally cannot do: be accountable

There's one complement that matters more than the rest, and it's the one a model can't supply no matter how good it gets. A machine that moves, one that can damage a part or hurt a person, needs someone who can be held responsible for it: a name on the work, someone who carries the reputational and legal weight, who can be looked in the eye when it matters. A model cannot be accountable. It can't sign off, can't be liable, can't stand behind a decision when a regulator or an insurer or an injured person asks who is answerable. The more capable the automation, the more that accountable human layer is worth. Not less.

Some ground commoditizes slower, on purpose

Not all software erodes at the same rate. In safety-critical and regulated work, anything where a mistake has physical or irreversible consequences, the very things that slow a project down elsewhere become the moat: liability, regulation, insurance, the hard requirement that a responsible person stand behind the system. That friction means the floor drops slowly, because "cheap and fast" was never the whole ask; "provably safe, and someone answerable" was. It's not a coincidence that the domains where verification and accountability dominate are the ones least threatened by abundant generation.

What this means for how I work

I'm not betting that I write code better than a model. That's a bet you lose. I'm betting on the layer above the code: taking a powerful capability and being the accountable, trusted person who makes it safe to hand a consequential system to. That's why the way I work leans on the things that survive abundance (proof over claims, verification built in, bounded consequences, honest scope) and why I gravitate to problems where being right and being answerable matter more than being fast. When building is free, that's where the value went.

← All notes