Who sees what online — and why

Andreas Haupt is a PhD candidate in Social and Engineering Systems (SES) at IDSS, and the Computer Science and Artificial Intelligence Laboratory (CSAIL). He obtained his MS in Economics and Mathematics from the University of Bonn, and his BS in Computer Science from the University of Frankfurt. He has worked in platform enforcement at the European Commission’s antitrust division and the Federal Trade Commission. Outside of work, he has led the MIT Science Policy Initiative and the MIT AI Ethics and Policy group.

What is the focus of your research? What sort of knowledge and disciplines does it bring together? How will it make an impact?

I study how the design of algorithms affects who sees what and why on online platforms. Recommendation systems choose content to show to users based on their previous engagement, such as in YouTube, Instagram, or TikTok feeds. The deployed algorithms predict how much users will “like” content, and show predicted-to-be-liked content. From this premise, three main challenges come in and inform my work: Does the algorithm characterize users in predicting (user models)? Is this even performing well (online learning)? And how does this influence what content appears on the platform in the first place (supply side)?

Platforms are becoming more important and shaped by advances in artificial intelligence. With more and more of our time spent online, and more informed ways for algorithms to decide what people want, platforms’ importance is, I believe, on the rise.

I am using methods from Artificial Intelligence, in particular Reinforcement Learning, and Economics, specifically from the fields of (Empirical) Industrial Organization and Microeconomic Theory. These are complementing each other: Microeconomic Theory provides initial intuitions for the study of platforms; Empirical Industrial Organization provides quantitative models after confronting the models with data; and Reinforcement Learning provides many of the algorithms that we may increasingly see in platforms, feeding back into better models for Microeconomic Theory, and completing the circle.

What did you do before IDSS and why did you choose SES?

I was a high school teacher placed through a German non-profit. I was teaching mathematics and computer science. I was happy with the connections I had, but was also ready for growth. Leaving Europe for the U.S. had already been a big move. I chose IDSS because it offers the opportunity to work with computer scientists and economists, and to become a disciplinary hybrid, something I’ve learned to appreciate more and more over time.

What are your favorite things about the IDSS and MIT communities? What do you do for fun?

Lending a term that was introduced for Steve Jobs, a Reality Distortion Field seems to encompass MIT. MIT makes me believe almost anything is possible. I could have not imagined the ambition I would develop while studying at MIT. The Reality Distortion Field is what I will miss the most about MIT. IDSS is special even in this, in that it considers the big problems and creates socio-technical engineers that can go end-to-end in areas of study of their choice.

My private interests are more mundane: I like cooking, chance encounters at the many events around Cambridge, and the feeling of accomplishment after an early gym session with my friends.