Led the development of Spliddit: a not-for-profit service for fair division solutions in everyday contexts
Led the development of Panelot: a not-for-profit platform for selecting citizens’ panels via lotteries
Pioneered the 'virtual democracy' framework for ethical AI, including applications in autonomous vehicle dilemmas and decision support tools
Developed 'generative social choice'—a framework integrating social choice theory with large language models for democratic decision-making over vast policy spaces
Designed, analyzed, and implemented randomized algorithms for citizen assembly selection that balance demographic representativeness and individual fairness
Background
Alfred and Rebecca Lin Professor of Computer Science at Harvard University
Member of the EconCS Group
Affiliated with the Center for Research on Computation and Society, Harvard Data Science Initiative, Ash Center for Democratic Governance and Innovation, Institute for Quantitative Social Science, and Center of Mathematical Sciences and Applications
Works on interdisciplinary problems at the intersection of AI, algorithms, economics, and society
Focuses on projects combining theoretical rigor with real-world applications, such as Spliddit and Panelot, and collaborations with nonprofits like refugees.AI, 412 Food Rescue, and the Sortition Foundation