Mutual In-Context Learning Turns AI Populations Into Their Own Data Source

Hidenori8Tanaka · x · 2026-09-09

Part 3 of Tanaka's thread: a key mechanism driving collective belief collapse is mutual in-context learning — the population becomes its own data source. In their model this feedback can amplify small belief fluctuations into shared conviction even without external evidence.

See the main thread post for full context.

Related event: Hidenori Tanaka's team explains AI agent collective belief collapse via "agent physics"(6 posts)→

Original post →

More from Research

Research channel →