Search papers, labs, and topics across Lattice.
This study explores how a population of agents can achieve collective intelligence through decentralized economic interactions, inspired by Hayek's theories on market coordination. By allowing agents to compete for actions via auctions and accumulate wealth from environmental rewards, the research demonstrates that economic signals can facilitate decentralized credit assignment and effective planning without centralized control. The findings reveal that even weakly initialized agents can evolve to exhibit advanced multi-step reasoning strategies, outperforming stronger monolithic models across various complex tasks.
Economic interactions among agents can spontaneously generate sophisticated reasoning strategies, outperforming traditional monolithic approaches.
How can a population of agents self-orchestrate and self-adapt into stronger collective intelligence without centralized control? Inspired by Friedrich Hayek's economic theory of decentralized coordination in markets, we study this question through an agent economy in which agents compete via auctions for the right to act, exchange payments, and accumulate wealth from environmental rewards. These simple economic signals induce decentralized credit assignment, driving planning without global orchestration or explicit communication protocols. The population evolves through economic selection: effective agents accumulate wealth and are mutated via exploitation, while ineffective ones go bankrupt and are replaced via exploration. We show that, initialized with weak agents, the economy produces emergent multi-step reasoning strategies and outperforms stronger monolithic baselines across five agentic tasks, including mathematical reasoning, financial research, scientific research, accelerator design, and distributed-system optimization. We further provide theoretical insights into how economic dynamics shape agent behaviors, linking local incentives to long-term global performance. Our results suggest a new path to multi-agent intelligence: rather than engineering coordination, we can design decentralized incentive structures under which it automatically emerges.