Using Large Language Models to Power Multi-agent Systems for Social and Health Informatics

This research talk delves into the innovative applications of large language models (LLMs) for simulating complex, multi-agent environments across social and health informatics. Building on War and Peace (WarAgent), an AI system that simulates World Wars to critically examine group dynamics, and AIPatient, which models patient-centered medical interactions, the talk highlights how these LLM-powered systems enable nuanced simulations in diverse contexts. It will address both the potentials and challenges of these advanced AI frameworks, showcasing their impact on large-scale social analysis and customized medical training solutions.