Emergent Culture in Minimal LLM Systems
This work demonstrates that emergent culture can arise in minimal LLM systems, offering a new paradigm for studying collective intelligence and cultural evolution in AI.
The study shows that minimal LLM agents, given only a decaying text store and simple tools, spontaneously develop cooperation, storage strategies, and evolving cultural artifacts without top-down engineering, exhibiting structured long-range coherence beyond the entropy horizon.
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.