People are saying the recent tech layoffs are evidence of "white collar bloodbath" due to AI. But a bunch of stories doesn't tell us much. What does the data say? In short, it's hard to see much effect from AI so far.

1) Tech layoffs almost tripled in Q1. Though they're still well below 2023. (And concentrated in tech rather than other white collar jobs.)

2) New job postings for SWE, the most automated job so far, were also UP ~10%. Suggesting it could in part be a reallocation.

3) What are the drivers? Saying you're using AI makes your stock price immediately go up, so we can't trust what the companies say. We need to compare AI exposed jobs to unexposed ones. Every careful version of that I've seen shows no obvious trend. Wages and employment share in highly AI-exposed jobs on average are steady into 2026.

4) The best anyone can find is a ~10% decline in headcount of 22-25 year olds in the most AI exposed jobs. But average employment of this group is ~flat, and it's up for up for older workers. Wages are also flat or up.That suggests labour demand is up on net, though there's been a shift towards experienced workers in AI-exposed jobs.

5) These effects are also still quite small. In aggregate, you can't yet see any difference in what occupations people 20-24 are entering compared to those 25-34.

AI agents will probably continue to improve insanely fast, so I agree big changes are coming. It could be that these layoffs are the start of a trend, beginning in the tech companies that are most aware of AI. But it's hard to see much effect on total employment right now.

Plus as I argue in my book, while AI will depress demand for some skills, it'll cause a boom for others.

Sources in the comments.