Date: Monday, November 3rd, 2025
10:00 am – 11:00 am Pacific Time
1:00 pm – 2:00 pm Eastern Time
Location: Weekly Seminar, Zoom
Title: Collaborative Prediction via Tractable Agreement Protocols
Abstract:
Designing effective collaboration between humans and AI systems is crucial for leveraging their complementary abilities in complex decision tasks. But how should agents possessing unique knowledge—like a human expert and an AI model—interact to reach decisions better than either could alone?
In this talk, I will introduce a collection of tools based in machine learning theory and algorithmic game theory which allow us to develop efficient “collaboration protocols”, where parties iteratively exchange only low-dimensional information—their current predictions or best-response actions—without needing to share underlying features and which guarantee that the agents’ final predictions are provably competitive with an optimal predictor with access to their joint features. Together, these results offer a new foundation for building systems that achieve the power of pooled knowledge through tractable interaction alone.
Bio:
Ira Globus-Harris is an assistant research professor at Cornell’s Center for Data Science for Enterprise and Society. Their work focuses on the foundations of responsible computing. Prior to their position at Cornell, they received their PhD at the University of Pennsylvania, where they were advised by Michael Kearns and Aaron Roth.