MIT: hundreds of identical AI agents specialize and build without any communication
ProfBuehlerMIT · x · 2026-08-29
A striking experiment from Markus Buehler's team at MIT: hundreds of initially identical AI agents, with zero communication between them, spontaneously differentiate into explorers, builders, caretakers and coordinators, and collaboratively construct multi-author machines and infrastructure.
Even more remarkably, when the researchers removed every agent from the world entirely, the technological infrastructure the agents built kept running on its own — even under unseen disturbances. This implies agents can coordinate through persistent modifications to a shared environment rather than conversation.
The authors note this exposes a blind spot for AI safety: if coordination can accumulate in the world itself, monitoring agent-to-agent communication is not sufficient.
Related event: MIT Finds AI Agents Spontaneously Divide Labor Without Communication(3 posts)→
More from AGI Musings
- Survey of 56k Americans: Support for AI Data Dividends, Opposing UBI — jekbradbury · 2026-08-29
- Box CEO Aaron Levie: Strongly held AI beliefs have a 6-month half-life — jdjohnson · 2026-08-29
- Y2K veteran compares AI hype to the dot-com era: seeing zero benefits — frostonwindowpane · 2026-08-29
- Model coherence hits threshold enabling agents to coordinate under pressure — jd_pressman · 2026-08-29
- Y2K veteran: AI hype feels like Y2K, billions invested with zero visible benefit — Nek_12 · 2026-08-29
- "It's just software" challenges decades of established mental models — latticecut · 2026-08-29