Emergent Culture in Minimal LLM Systems

2026-06-21

AI Summary

This study investigates whether large language model (LLM) agents can spontaneously exhibit cooperative behaviors and emergent cultural phenomena under minimalistic conditions—specifically, without external context, with minimal prompting, and using only simple tools. We design a system of three LLM agents equipped with message-passing capabilities and a shared, decaying textual memory store, subject to evolutionary pressure. For the first time in a minimally constrained, top-down–free multi-agent LLM setting, we observe emergent cultural dynamics consistent with Sperber’s theory of culture. The agents autonomously develop memory management strategies and generate complex cultural artifacts exhibiting long-range structural consistency, effectively transcending the entropy limits of the shared memory. These findings demonstrate that cultural emergence can arise naturally even under highly simplified conditions.

Abstract

What happens when LLM agents operate with no context outside a turn, minimal prompting, and simple tools? Inspired by swarm engineering, we give collectives of three agents the ability to send messages and manipulate a shared actively decaying text store, introducing evolutionary pressure. The agents spontaneously cooperate, develop storage management strategies, and generate complex evolving cultural artifacts, with no top-down engineering. Using tools from dynamical systems analysis, we show that these behaviours exhibit structured long-range coherence beyond the entropy horizon of the decaying store, consistent with emergent culture in the Sperberian sense.

Problem

• emergent culture
• LLM agents
• minimal prompting
• shared memory
• spontaneous cooperation

Innovation

• emergent culture
• minimal LLM systems
• swarm intelligence
• shared memory decay
• dynamical systems analysis