Super-additive Cooperation in Language Model Agents

TL;DR

This paper demonstrates that intergroup rivalry and repeated interactions among language model agents in a simulated Prisoner's Dilemma significantly enhance cooperation, providing insights for designing cooperative multi-agent AI systems.

Contribution

It introduces a novel framework showing how super-additive effects of competition and repetition promote cooperation in language model agents.

Findings

Intergroup rivalry increases cooperation levels.

Repeated interactions boost initial cooperation.

The framework models complex social behaviors in AI agents.

Abstract

With the prospect of autonomous artificial intelligence (AI) agents, studying their tendency for cooperative behavior becomes an increasingly relevant topic. This study is inspired by the super-additive cooperation theory, where the combined effects of repeated interactions and inter-group rivalry have been argued to be the cause for cooperative tendencies found in humans. We devised a virtual tournament where language model agents, grouped into teams, face each other in a Prisoner's Dilemma game. By simulating both internal team dynamics and external competition, we discovered that this blend substantially boosts both overall and initial, one-shot cooperation levels (the tendency to cooperate in one-off interactions). This research provides a novel framework for large language models to strategize and act in complex social scenarios and offers evidence for how intergroup competition can, counter-intuitively, result in more cooperative behavior. These insights are crucial for designing future multi-agent AI systems that can effectively work together and better align with human values. Source code is available at https://github.com/pippot/Superadditive-cooperation-LLMs.