AImpact Center Seminar Series: Dr. Alex Moehring, "Designing Human-AI Collaboration: A Sufficient Statistic Approach."

Abstract: We develop a sufficient-statistic approach to designing collaborative human-AI decision-making policies in classification problems, where AI predictions can be used to either automate decisions or selectively assist humans.

The approach allows for endogenous and biased beliefs, and effort crowd-out, without imposing a structural model of human decision-making. We deploy and validate our approach in an online fact-checking experiment.

We find that humans under-respond to AI predictions and reduce effort when presented with confident AI predictions. AI under-response stems more from human overconfidence in own-signal precision than from under-confidence in AI.

The optimal policy automates cases where the AI is confident and delegates uncertain cases to humans while fully disclosing the AI prediction. Although automation is valuable, the additional benefit from assisting humans with AI predictions is negligible.

Bio: Alex is an assistant professor at the Daniels School of Business at Purdue University. His research focuses on the economics of digitization with an emphasis on digital platforms, Human-AI interaction, and social influence.

Alex holds a PhD from the MIT Sloan School of Management in addition to a BS in Business Administration and BA in Economics from the University of North Carolina at Chapel Hill. He has also worked as a research assistant at Microsoft Research and an analyst at Wells Fargo Securities.