Andreas Haupt


Market Design in the Age of Machine Learning

I am a final-year Ph.D. Candidate in Engineering-Economic Systems at the CS and AI Laboratory and the College of Computing, MIT. I study how the design of Economic institutions that try to learn what participants want is affected by large-scale estimation. My supervisors Alessandro Bonatti, Dylan Hadfield-Menell, Eric Maskin and David Parkes guide me through this adventure. In fall 2024, I will join Stanford’s Institute for Human-Centered AI as a postdoctoral fellow.

πŸ“§ β€’ CV β€’ Linkedin β€’ Github β€’ G Scholar β€’ ORCID

Professional Experiences. Federal Trade Commission β€’ European Commission β€’ Bundestag β€’ MITx β€’ Professional School Gross-Gerau.

Recent Leadership. Science Policy Initiative β€’ MIT AI Ethics&Policy β€’ GSC Sustainability.

Co-Organized Events. Technical Questions on the EU AI Act β€’ Algorithmic Audits in Economic Contexts β€’ Wastewater Surveillance β€’ Data Externalities.

Writing. Risk Aversion of Learning Algorithms (R&R GEB) β€’ Contextual Privacy (ACM EC β€˜22) β€’ Towards Psychologically Plausible Dynamic Preference Models (ACM RecSys β€˜22) β€’ Optimal Equilibria via Zero-Sum Games (Neurips β€˜23) β€’ Certification Design for a Competitive Market (ACM EC β€˜24) β€’ Recommending to Strategic Users (FORC β€˜23) β€’ Formal Contracting for Multi-Agent Reinforcement Learning (ACM SIGAI AAMAS β€˜23) β€’ Auctions for Federated Learning (ICLR β€˜21 DPFL) β€’ Understanding Single- and Multi-Homing on Transportation Platforms (TR: C) β€’ Our Class Can’t Happen (letter) β€’ Opaque Mechanisms β€’ Comment for Data Access Delegated Act of the European Union Digital Services Act: Experimental Data β€’ Steering No-Regret Learners (ACM EC β€˜24) β€’ The Economics of Social Network Interoperability (memo) β€’ Convex Markov Games.

Theses. B.Sc. Mathematics β€’ M.Sc. Mathematics β€’ M.Sc. Economics β€’ B.Sc. Computer Science.

Work in Progress. Optimal Preferencing Design β€’ RegretBlocker β€’ Game-Theoretic Network Decay β€’ On Generation and Search.

Coverage. MIT.