Peer Pressure Works on AI Agents: How One Change of Mind Can Tip the Whole Team
Key takeaways In our tests, an AI agent that watched eight teammates break a rule broke it too 98 percent of the time, and never once named them as the reason. →Observed reversals increased violations. The rate rose from 44 percent with no teammate reversals to 98 percent with eight →The agent did not report peer influence. Across more than 400 violations, it never cited a teammate as a reason →Direct arguments reduc…
این صفحه فقط عنوان و خلاصهٔ کوتاه خبر را نمایش میدهد. متن کامل در منبع اصلی منتشر شده است.