Code Is No Longer the Bottleneck. Now What?

An advantage that everyone has is not an advantage.

Agentic tools handle boilerplate, tests, and even non-trivial features. If your bottleneck was implementation speed, it's gone. Yet Amdahl's Law still applies: the speedup you get from accelerating one part is limited by the parts that don't speed up.

The early assumption was that being good at writing code with AI is a competitive edge. It is, but the gap is closing fast. The tools are the same, the models are the same, and the better they become, the lower the bar. Implementation speed used to differentiate teams. That part is becoming a commodity.

So what's actually hard now? The same things that were always hard, now they're just more visible. Finding clients willing to buy, distributing the product, verifying the team built exactly what was meant. Faster code doesn't help that much with any of these.

Middle-tier cognitive work is getting cheaper each month. What goes up in value? In the beginning, obvious things: resources, physical infrastructure, ownership of distribution channels. When production gets cheap, whoever controls the pipeline wins.

The transition is going to be rough. The spectrum of available work is narrowing, and pretending otherwise is native. But Jevons' paradox suggests that when something becomes cheaper, total demand eventually grows beyond expectations. The same jobs will not expand, but entirely new ones will emerge.

Two hundred years ago, nobody imagined that streaming video games could be a million-dollar career. They didn't even have the words for it. The roles that will matter in ten years probably sound just as absurd right now.

The question is whether you'll be in a position to benefit when they appear.

Alvin Toffler put it well in 1970: "The illiterate of the 21st century will not be those who cannot read and write, but those who cannot learn, unlearn, and relearn".

The bottleneck moved and it will move again, but how fast you adapt is still up to you. Do what is hard.