AI for Social Impact: Deployed Resource Optimization and Future Acceleration with Foundation Models

Abstract: For nearly two decades, my team’s work on AI for Social Impact (AI4SI) has focused on optimizing limited resources in critical areas like public health, conservation, and public safety. I will highlight recent results from our deployed work in India, demonstrating measurable improvements in effectiveness for the world's two largest mobile health programs for maternal and child care, which have served millions of beneficiaries. We have leveraged innovative restless and collaborative bandit algorithms to achieve these gains, revealing new technical directions in the process. Additionally, I will touch upon our previous work on influence maximization for HIV prevention among youth experiencing homelessness in Los Angeles. Deploying end-to-end AI4SI systems requires us to repeat three essential steps: understanding stakeholders’ resource allocation challenges, building a tailored model, and rigorously testing in the field. I'll share initial results on how we can leverage foundation models and LLMs to accelerate this AI4SI process, potentially revolutionizing the speed and scale of social impact applications that focus on resource optimization.

Bio: Milind Tambe is Gordon McKay Professor of Computer Science and Director of Center for Research on Computation and Society at Harvard University; concurrently, he is also Principal Scientist and Director for "AI for Social Good" at Google Deepmind.  Prof. Tambe and his team have developed pioneering AI systems that deliver real-world impact in public health (e.g., maternal and child health), public safety, and wildlife conservation. He is recipient of the AAAI  Award for Artificial Intelligence for the Benefit of Humanity,    AAAI Feigenbaum Prize,   IJCAI John McCarthy Award,    AAAI Robert S. Engelmore Memorial Lecture Award, AAMAS ACM Autonomous Agents Research Award,  INFORMS  Wagner prize for excellence in Operations Research practice, Military Operations Research Society Rist Prize,  Columbus Fellowship Foundation Homeland security award and commendations and certificates of appreciation from the US Coast Guard, the Federal Air Marshals Service and airport police at the city of Los Angeles.  He is a fellow of AAAI and ACM.

From Click to Delivery: Challenges and Opportunities in Multi-Agent Path Finding at Amazon

Abstract: Amazon tackles the problem of coordinating hundreds of thousands of robots to go from a website click to a delivery at your door. We leverage problem, heuristic search, and ML to solve this huge optimization problem. In this talk I will provide an overview of the underlying motion, allocation, perception, and manipulation sub-problems, and discuss some of the challenges and opportunities their integration poses for research in MAPF.

Bio: Federico Pecora leads the Movement Science team at Amazon Robotics. His team studies algorithms for efficient motion planning and coordination, with the aim of improving the efficiency the world’s largest fleet of mobile robots. He leads Amazon Robotics’ effort to develop a Foundation Model for structured field mobile robots. Prior to joining Amazon, Federico was head of the Multi-Robot Planning and Control Lab at Örebro University, Sweden, where he was also professor and director of the Computer Science Engineering program.

On AI Alignment in the Provision of Social Services: Opportunities and Challenges

Abstract: Artificial intelligence and machine learning are increasingly used to aid decision-making about the allocation of scarce societal resources, for example housing for homeless people, organs for transplantation, and educational supports for K-12 students. What does it mean for these systems to be "aligned" with human preferences? In practice, this involves attempts to achieve some combination of fairness, efficiency, and incentive compatibility, depending on the preferences of some set of stakeholders. In this talk I will give an overview of my group's research in this space, informed by the theories of local justice and of street level bureaucracy. I will discuss our work on characterizing and analyzing the efficiency, the fairness, and the distributive justice implications of human, machine, and human+machine decision-making in public service provision, with a particular focus on resources that serve those experiencing, or at high risk of, homelessness. I will give a peak into theoretical, empirical, and experimental results, and discuss where I think AI can be most helpful, as well as significant human and technical challenges.

Bio: Sanmay Das is Professor of Computer Science and Associate Director of AI for Social Impact at the Sanghani Center for AI and Data Analytics at Virginia Tech’s Innovation Campus. He is chair of the ACM Special Interest Group on Artificial Intelligence, a member of the DARPA ISAT Study Group, and an emeritus member of the board of directors of IFAAMAS. He serves as an associate editor for ACM TEAC, JAIR, and JAAMAS. He has served as program co-chair of AAMAS and of AIES, and as Associate Program Chair for IJCAI. He has been recognized with awards for research, teaching, and service, including a National Science Foundation CAREER Award, the Department Chair Award for Outstanding Teaching at Washington University, and the Outstanding Service Award from the Computer Science Department at George Mason University. He was selected as an ACM Distinguished Member in 2023 for contributions to AI and economics, AI for social good, and for service to the profession.