SMU SOE Seminar Series (August 6, 2025): Calibrated Coarsening: Designing Information for AI-Assisted Decisions

TOPIC:

ABSTRACT

Artificial intelligence (AI) signals are increasingly deployed as human decision-making aids across many critical applications, but human cognitive biases can prevent them from improving outcomes. We propose calibrated coarsening—partitioning the signal space into fewer cells at optimised thresholds—as a way to improve decision-making outcomes while (i) keeping humans in the loop, (ii) modifying signals without deception, and (iii) adapting flexibly to various cognitive biases and decision-making contexts. Within an optimal information disclosure framework, we derive the approximately-optimal universal coarsened policy for settings where the designer does not observe the decision-maker’s information. We then empirically demonstrate in a randomised experiment involving loan specialists that coarsening AI signals at the theory-derived threshold significantly improves decision-making outcomes, over both the human-only (based solely on the loan application) and continuous AI (assisted with uncoarsened AI risk-score) benchmarks. We uncover substantial decision heterogeneity amongst loan officers, and use a Bayesian hierarchical model to personalise coarsening policies, which can further improve outcomes as past data become available.

Click here to view the CV.
Click here to view the paper.