US programmer job growth nearly halved since ChatGPT launched, Fed study finds

The Fed study is the first government-level quantification of AI's displacement effect on a white-collar technical role, which matters because it shifts the conversation from anecdote and survey data to official labor statistics. Critically, the halving of growth is not the same as net job losses — the baseline trend and counterfactual matter enormously for interpreting severity.

The TechCrunch piece from April 18 on the App Store boom is the most relevant thread here: if AI tools are lowering barriers to app creation and driving a surge in new mobile software, that suggests output is rising even as the headcount producing it contracts. More developers shipping more apps with fewer programmers is exactly the productivity-displacement dynamic the Fed data would predict. The Cerebras IPO coverage from the same week is relevant in a different register — capital is flowing heavily into AI infrastructure, which tells you where the industry expects value to accumulate: in the model and chip layer, not in the labor that writes the code those models increasingly assist or replace.

Watch whether the Bureau of Labor Statistics' next Occupational Employment and Wage Statistics release, due in late 2026, corroborates the Fed's trend line across adjacent roles like QA engineers and data analysts — if those categories show similar deceleration, the effect is broad; if it's isolated to general programmers, the story is more about commoditization of entry-level coding specifically.