The apocalypse got benched
Remember that viral Citrini Research thought experiment about AI blowing unemployment out to 10.2% by 2028 and taking the S&P 500 down with it? Yeah, an economist just walked onto the field and said: not so fast. Julius Probst of Recruitonomics called that scenario “extremely unrealistic,” arguing that a jobless rate that high would basically light the economy on fire fast enough to trigger stimulus, rate cuts, and a very different policy response.
AI is eating some jobs, not all the jobs
Probst isn’t saying AI is harmless — he’s just saying the labor market isn’t one giant domino chain waiting to tip over. His split-screen version looks more like this:
- AI-skilled workers are getting a boost, with job postings that mention AI skills roughly tripling over the past couple of years.
- Routine white-collar roles like customer service, sales, and entry-level finance are the vulnerable ones.
- Blue-collar and physical jobs may actually be in a decent spot, especially with construction demand rising around the data-center buildout.
That last bit is the part investors should care about. If AI spending keeps turning into concrete, steel, and labor hours, the benefits ripple far beyond semiconductors and hyperscalers.
The Fed side quest
This all lands right in the middle of a sticky macro setup: payrolls are still growing, unemployment is holding up, and inflation is running above target. Probst argues the Fed may be being too chill about it, especially with the central bank meeting on June 16 and June 17.
So the big question isn’t just whether AI changes jobs over time. It’s whether the market has been treating AI like a clean productivity story when, in reality, it’s also a giant demand engine for labor, power, and capital spending.
Big picture: the AI trade may be less “robots steal all the jobs” and more “some jobs shift, some wages rise, and the Fed still has to deal with sticky inflation.” And that’s a very different movie.