AI-Assisted Decision Making with Human Learning

TL;DR

This paper explores how AI systems can optimally select features to support human decision-making, balancing immediate prediction accuracy with long-term human learning, influenced by the algorithm's patience and human learning ability.

Contribution

It introduces a framework analyzing the tradeoff between educating humans and optimizing short-term predictions in AI-assisted decisions, with a tractable characterization of optimal feature selection.

Findings

Optimal feature selection depends on the algorithm's patience and human learning capacity.

More patient algorithms choose more informative features, improving accuracy and understanding.

The optimal feature subset sequence is computationally tractable to determine.

Abstract

AI systems increasingly support human decision-making. In many cases, despite the algorithm's superior performance, the final decision remains in human hands. For example, an AI may assist doctors in determining which diagnostic tests to run, but the doctor ultimately makes the diagnosis. This paper studies such AI-assisted decision-making settings, where the human learns through repeated interactions with the algorithm. In our framework, the algorithm -- designed to maximize decision accuracy according to its own model -- determines which features the human can consider. The human then makes a prediction based on their own less accurate model. We observe that the discrepancy between the algorithm's model and the human's model creates a fundamental tradeoff: Should the algorithm prioritize recommending more informative features, encouraging the human to learn their importance, even if it results in less accurate predictions in the short term until learning occurs? Or is it preferable to forgo educating the human and instead select features that align more closely with their existing understanding, minimizing the immediate cost of learning? Our analysis reveals how this trade-off is shaped by both the algorithm's patience (the time-discount rate of its objective over multiple periods) and the human's willingness and ability to learn. We show that optimal feature selection has a surprisingly clean combinatorial characterization, reducible to a stationary sequence of feature subsets that is tractable to compute. As the algorithm becomes more "patient" or the human's learning improves, the algorithm increasingly selects more informative features, enhancing both prediction accuracy and the human's understanding.