Working Papers
November 2025 - R&R, Review of Economic Studies
with David Autor, Andrew Caplin, and Philip Marx
We demonstrate the failure of the aligned learning premise in two focal machine learning applications and rationalize this failure using an economic model of incentive design with endogenous information acquisition.
Humans and AI / Attention and Perception / Information Disclosure
April 2026
Learning from an Unknown DGP: Experimental Evidence on Belief Updating with AI Recommendations
with Matthew Kovach and Gerelt Tserenjigmid
We experimentally study belief updating when a decision maker does not know the DGP, but learns from qualitative statements such as AI recommendations.
Humans and AI / Attention and Perception / Information Disclosure
April 2026 - Submitted
with David Almog, Romain Gauriot, and Lionel Page
We provide field evidence that AI oversight can affect human decision-making by investigating Hawk-Eye review of umpires in top tennis tournaments.
Humans and AI / Attention and Perception
January 2026 - Submitted
with David Almog and Lucas Lippman
We use an experiment with a real work task to study whether workers change behavior when they know AI will be used to judge their work instead of humans.
Humans and AI / Attention and Perception / Information Disclosure
February 2026 - R&R, Behavior Research Methods
Improving Crowdsourcing for AI through Cognitive-Inspired Data Engineering
with Gunnar Epping, Andrew Caplin, Erik Duhaime, William Holmes, and Jennifer Trueblood
We investigate whether ideas from cognitive science can mitigate cognitive constraints and biases in crowdsourced datasets and improve models trained on these data.
Humans and AI / Attention and Perception
March 2026 - Submitted
Managing Cognitive Bias in Human Labeling Operations for Rare-Event AI: Evidence from a Field Experiment
with Gunnar Epping, Andrew Caplin, Erik Duhaime, William Holmes, and Jennifer Trueblood
We run a field experiment on DiagnosUs to address the prevalence effect in the AI lifecycle.
Humans and AI / Attention and Perception
Published Papers
Management Science, 2025
with Andrew Caplin, David Deming, Shangwen Li, Philip Marx, Ben Weidmann, and Kadachi Jiada Ye
We use a controlled experiment to show that ability and belief calibration jointly determine the benefits of working with AI.
Humans and AI / Attention and Perception
Journal of Economic Theory, 2025
with Andrew Caplin and Philip Marx
We investigate whether the predictions of modern machine learning algorithms are consistent with economic models of human cognition.
Humans and AI / Attention and Perception
PLoS ONE, 2024
with Nir Chemaya
We survey academics about norms around reporting ChatGPT use in manuscript preparation and test GPT-modified abstracts with AI detection software.
Humans and AI / Information Disclosure
Decision Analysis, 2026
with Gunnar Epping, Andrew Caplin, Erik Duhaime, William Holmes, and Jennifer Trueblood
We propose an approach to AI-augmented decision-making systems that uses human labeler uncertainty to improve accuracy and alignment with expert uncertainty.
Humans and AI / Attention and Perception