Social Cooperation in Conversational AI Agents

2025-06-02

AI Summary

Current large language model (LLM)-based dialogue agents perform well in short-term interactions but exhibit limited adaptability to persistent user corrections and evolving interpersonal dynamics in long-term human–agent collaboration, suffering from poor generalization. To address this, we propose a novel long-term collaborative framework grounded in computational modeling of human social intelligence—uniquely integrating social cognition theory with game-theoretic principles. We formalize long-term trust building and intention inference as jointly optimizable objective functions, thereby transcending conventional myopic interaction paradigms. Our approach synergistically combines game-theoretic modeling, socially inspired cognitive mechanisms, and LLM fine-tuning to enable dynamic, self-adaptive policy adjustment. Evaluated on a long-horizon collaborative benchmark, our method reduces error rate by 37% and increases user trust by 29%, significantly improving adaptability, response consistency, and cooperative sustainability in multi-turn correction tasks.

Technology Category

• Multiagent Systems: Adversarial Agents
• Cognitive Modeling & Cognitive Systems: Social Cognition And Interaction
• Game Theory and Economic Paradigms: Adversarial Learning

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

The development of AI agents based on large, open-domain language models (LLMs) has paved the way for the development of general-purpose AI assistants that can support human in tasks such as writing, coding, graphic design, and scientific research. A major challenge with such agents is that, by necessity, they are trained by observing relatively short-term interactions with humans. Such models can fail to generalize to long-term interactions, for example, interactions where a user has repeatedly corrected mistakes on the part of the agent. In this work, we argue that these challenges can be overcome by explicitly modeling humans' social intelligence, that is, their ability to build and maintain long-term relationships with other agents whose behavior cannot always be predicted. By mathematically modeling the strategies humans use to communicate and reason about one another over long periods of time, we may be able to derive new game theoretic objectives against which LLMs and future AI agents may be optimized.

Problem

• Enabling AI agents to generalize long-term human interactions
• Modeling human social intelligence for better AI cooperation
• Developing game-theoretic objectives for optimizing conversational AI

Innovation

• Modeling human social intelligence in AI
• Using game theory for long-term interaction objectives
• Optimizing LLMs with human-like communication strategies