Towards Foundation-model-based Multiagent System to Accelerate AI for Social Impact
Yunfan Zhao, Niclas Boehmer, Aparna Taneja, Milind Tambe
TL;DR
The paper tackles the labor-intensive nature of AI4SI development by proposing a meta-level multi-agent system that uses foundation models to accelerate problem formulation, solution design, and testing for resource-allocation tasks. FM-agents configure and adapt base-level agents and methods across scenarios, with human-in-the-loop oversight and fairness constraints. The framework aims to lower costs, improve generalization across domains, and enable safer deployment in social-impact settings, illustrated through the ARMMAN running example for designing state-action spaces, adaptive solution methods, and fairness-aware deployment.
Abstract
AI for social impact (AI4SI) offers significant potential for addressing complex societal challenges in areas such as public health, agriculture, education, conservation, and public safety. However, existing AI4SI research is often labor-intensive and resource-demanding, limiting its accessibility and scalability; the standard approach is to design a (base-level) system tailored to a specific AI4SI problem. We propose the development of a novel meta-level multi-agent system designed to accelerate the development of such base-level systems, thereby reducing the computational cost and the burden on social impact domain experts and AI researchers. Leveraging advancements in foundation models and large language models, our proposed approach focuses on resource allocation problems providing help across the full AI4SI pipeline from problem formulation over solution design to impact evaluation. We highlight the ethical considerations and challenges inherent in deploying such systems and emphasize the importance of a human-in-the-loop approach to ensure the responsible and effective application of AI systems.