cooperation

This paper examines how persona prompts influence strategic behavior of large language model agents in an iterated Split or Steal game, finding that mutual Split outcomes dominate and that model choice and persona type significantly affect cooperation and exploitation.

Article reports on a Beijing AI conference where top experts from the US and China expressed alarm over the risks of advanced AI and called for international cooperation to mitigate cybersecurity and systemic dangers, drawing parallels to Cold War nuclear coordination.

This paper argues that superintelligent AI systems designed under a solipsistic paradigm that treats the world as stationary will be self-undermining and uncooperative, leading to collective failures. The authors call for a new research paradigm that treats interdependence and cooperation as core design principles.

Introduces SovSim, a multi-agent simulation framework for studying cooperation and resource sustainability in LLM societies with asymmetric power structures. Experiments show that introducing a dominant agent (boss or king) severely degrades cooperation and survival rates across 11 state-of-the-art models.

This paper presents a two-level autoresearch framework where an outer-loop AI agent autonomously optimizes inner-loop LLM policy-synthesis pipelines for multi-agent sequential social dilemmas, achieving superior performance and discovering objective-specific mechanisms like fairness under a maximin welfare objective.

OpenAI presents LOLA (Learning with Opponent-Learning Awareness), a multi-agent reinforcement learning method where agents shape the anticipated learning of other agents. The approach demonstrates emergence of cooperation in iterated prisoner's dilemma and convergence to Nash equilibrium in game-theoretic settings.