AI Companies Promised Unemployment and Now Nobody Wants to Hear It Was a Mistake

A Bluesky user put it plainly this week, and twelve people liked it, which in Bluesky terms means it landed in the right rooms: if the AI companies didn't want a backlash, they shouldn't have spent the last two years on a press tour warning that AI would cause 20% unemployment or higher and acting like that was fine. The post wasn't a hot take — it was an accounting. The industry spent years building public appetite for its own consequences, and now those consequences are arriving in the form of actual layoffs, and the same industry is surprised that people are angry.

A second post, circulating in the same current, asked a question that nobody in a boardroom seems to have a clean answer to. If AI is boosting productivity, the argument went, why are companies shrinking instead of growing? The analogy was agricultural: if you suddenly had two tractors, you'd work two fields, not fire half your farmhands. The post framed the layoffs not as inevitable technological progress but as a choice dressed up as inevitability — companies using AI as a socially acceptable excuse to do what they wanted to do anyway. The confusion it expressed wasn't naive. It was the kind of confusion that comes from watching an argument collapse under its own logic.

This is the specific tension driving the job displacement conversation right now. The industry's rhetorical strategy was always slightly incoherent: pitch AI to investors as a productivity multiplier that justifies enormous valuations, pitch it to regulators as an unstoppable force requiring accommodation, and pitch it to workers as a tool that augments rather than replaces. Those three pitches cannot all be true simultaneously, and workers are increasingly pointing that out. Microsoft's name appeared in roughly a fifth of recent posts on this beat — the company has become a kind of shorthand for the contradiction, a place where AI investment and workforce reduction have happened in the same fiscal year with the same press release energy. The optimists arguing that AI job loss is overhyped are still getting traction, but they're swimming against a tide that the industry itself helped create.

The mood among workers — and the people watching workers — has shifted in ways that make the standard reassurances sound hollow. The argument that AI creates more jobs than it destroys was always a historical claim about previous technological transitions, not a guarantee about this one. What the Bluesky posts this week captured is something simpler and harder to rebut: the people who told you to be afraid were the same people building the thing, and now that you're afraid, they want credit for your fear and absolution for the consequences. That's not a position that survives contact with someone who just lost their job.