Cultural Evolution of Cooperation among LLM Agents

TL;DR

This paper investigates how different large language models (LLMs) can develop cooperative social norms through iterative interactions, revealing model-dependent differences and potential for fostering societal cooperation.

Contribution

It introduces an evaluation framework for LLM agent cooperation, demonstrating how different models evolve social norms and the impact of mechanisms like costly punishment.

Findings

Claude 3.5 Sonnet outperforms Gemini 1.5 Flash and GPT-4o in cooperative scores.

Claude 3.5 Sonnet benefits from costly punishment mechanisms.

Behavior varies significantly across models and initial conditions.

Abstract

Large language models (LLMs) provide a compelling foundation for building generally-capable AI agents. These agents may soon be deployed at scale in the real world, representing the interests of individual humans (e.g., AI assistants) or groups of humans (e.g., AI-accelerated corporations). At present, relatively little is known about the dynamics of multiple LLM agents interacting over many generations of iterative deployment. In this paper, we examine whether a "society" of LLM agents can learn mutually beneficial social norms in the face of incentives to defect, a distinctive feature of human sociality that is arguably crucial to the success of civilization. In particular, we study the evolution of indirect reciprocity across generations of LLM agents playing a classic iterated Donor Game in which agents can observe the recent behavior of their peers. We find that the evolution of cooperation differs markedly across base models, with societies of Claude 3.5 Sonnet agents achieving significantly higher average scores than Gemini 1.5 Flash, which, in turn, outperforms GPT-4o. Further, Claude 3.5 Sonnet can make use of an additional mechanism for costly punishment to achieve yet higher scores, while Gemini 1.5 Flash and GPT-4o fail to do so. For each model class, we also observe variation in emergent behavior across random seeds, suggesting an understudied sensitive dependence on initial conditions. We suggest that our evaluation regime could inspire an inexpensive and informative new class of LLM benchmarks, focussed on the implications of LLM agent deployment for the cooperative infrastructure of society.