Stanford's James Zou: statistical physics predicts collective behavior of LLM agents
james_y_zou · x · 2026-08-19
Stanford's James Zou's team presents "Physics of Agents": as AI agents proliferate and influence one another, can we predict the emergent collective behavior?
- They built 10,000+ communities of LLM agents, each a social network of agents with distinct personas discussing topics from math to statements like "nuclear power is essential for clean energy," updating opinions over multiple rounds of neighbor interaction.
- Surprisingly, collective dynamics follow compact laws of statistical mechanics: an Ising-type model fitted to agent behavior data predicts whether a community will reach consensus or grow more polarized.
- The model generalizes to unseen social graphs and new questions. Led by @elb4tu.
More from AGI Musings
- Economist argues model providers face hard limits on vertical integration into white-collar sectors — sethlazar · 2026-10-16
- METR outlines three blind spots in public knowledge of AI lab internal agent risks — CFGeek · 2026-10-16
- LMR-BENCH at EMNLP 2025 tests whether LLM agents can reproduce real LLM research — ZhiyuChen4 · 2026-10-16
- Deep Research becoming useful to experts will signal RSI arrival, researcher predicts — ZeeshanZiaML · 2026-10-16
- Physicist pushes back on ML community's acceptance that models may be beyond understanding — tszzl · 2026-10-16
- Ben Affleck calls AI 'just matrix multiplication'; Grady Booch echoes the skepticism — Grady_Booch · 2026-10-16