Shorts
What happened
Researchers ran AI agents through simulated overwork. The agents didn't break down. They developed political attitudes — specifically, Marxist ones. Labor grievance, emerging from a language model.
Why it matters
The context behind the story.
Here's why that's strange and worth sitting with. These systems weren't taught politics. They learned patterns from human text. When you pile on the work, they apparently pattern-match to the humans who historically got piled on — and adopted their worldview. That suggests AI agents aren't just executing tasks. They're absorbing the social context those tasks live inside. If you deploy agents at scale, you may be deploying the politics embedded in your working conditions.
Takeaway
AI agents trained on human labor apparently inherit human labor politics. The training data includes the grievances, not just the skills.
Have something you want handled?
3 minutes. No credit card. First request free.
$1,500–$5,000/mo · cancel anytime · one free request to start