The Constraint Has Moved

Every time software development gets easier, the instinct is to assume there will be fewer developers. Scripting languages replaced assembly work. Higher-level frameworks replaced lower-level plumbing. Each shift looked like compression from the outside. And each time, the developer market expanded.

The reason is structural. When something gets cheaper to produce, consumption increases, not decreases. Economists call this Jevons' paradox: efficiency gains drive demand up. Software becomes faster to build, so more software gets built. Projects that weren't viable at the old price become viable at the new one. The floor doesn't drop. The ceiling rises.

The Job Isn't at Risk. Your Value Proposition Is.

The developer who fears AI replacing them has misread the threat. Writing code stops being a differentiator when the machine can write it faster. Boilerplate, standard integrations, predictable CRUD endpoints -- none of these are markers of professional value when the machine generates them on demand.

What opens up is everything that never got built. The internal tool that always got deprioritised. The feature backlog that never got touched. The project too expensive to justify last quarter. It all becomes viable now. And someone has to decide what to build, for whom, and at what cost to what else.

From Implementation to Judgement

The bottleneck used to be: can we build this? Now it's: should we build this? That question doesn't have a syntax. It doesn't autocomplete. It requires understanding the business, the users, the tradeoffs -- things that exist outside the codebase.

Most developers haven't started preparing for that shift. The job title hasn't changed yet, so the urgency doesn't feel real. But the version of the role that runs on execution is being commoditised. The version that runs on judgement is becoming more valuable by the month.

Figure out which one you're building.