The AI jobs narrative runs in one direction: the technology will displace workers. The less-covered story is what is happening to hiring right now, in the companies building the technology.
Across the tech sector, headcount growth has slowed sharply relative to the pre-2022 era — and not primarily because of the interest rate cycle. A significant portion of the slowdown reflects a deliberate bet by engineering leadership that AI tooling will allow existing teams to scale output without proportional headcount increases. The thesis is being tested in real time at major software companies, and the early returns are mixed but not unfavorable enough to reverse the bet.
The roles most visibly affected are mid-level and entry-level software engineering. The pipeline from boot camp and computer science degree to first job has compressed significantly. Not because those roles have been automated away, but because the productivity of senior engineers using AI tools has increased enough that companies need fewer junior engineers to move the same volume of work. The entry point is narrowing.
This creates a structural problem for the sector’s talent pipeline that is not yet showing up in unemployment statistics but is visible in hiring data, compensation surveys, and the quiet desperation of new graduates who expected a market that no longer exists in the form they planned for.
The long-term picture may normalize as software demand expands to absorb the productivity gains. That argument has historical precedent — technology tools have repeatedly increased demand for the labor that uses them. But the short to medium term is a genuine dislocation, and the people experiencing it are not well-served by the long-term narrative.
The disruption is already here. It just started in places the disruption story did not predict.