Human Expertise in Algorithmic Prediction

TL;DR

This paper presents a new framework for integrating human expertise into algorithmic predictions by identifying instances where human judgment can enhance algorithm performance, especially in cases where algorithms struggle.

Contribution

The paper introduces a novel method to incorporate human feedback based on algorithmic indistinguishability, improving prediction accuracy and enabling better human-AI collaboration.

Findings

Algorithms outperform humans on average.

Human judgment improves predictions on specific instances.

Nearly 30% of patients benefit from human input in X-ray classification.

Abstract

We introduce a novel framework for incorporating human expertise into algorithmic predictions. Our approach leverages human judgment to distinguish inputs which are algorithmically indistinguishable, or "look the same" to predictive algorithms. We argue that this framing clarifies the problem of human-AI collaboration in prediction tasks, as experts often form judgments by drawing on information which is not encoded in an algorithm's training data. Algorithmic indistinguishability yields a natural test for assessing whether experts incorporate this kind of "side information", and further provides a simple but principled method for selectively incorporating human feedback into algorithmic predictions. We show that this method provably improves the performance of any feasible algorithmic predictor and precisely quantify this improvement. We find empirically that although algorithms often outperform their human counterparts on average, human judgment can improve algorithmic predictions on specific instances (which can be identified ex-ante). In an X-ray classification task, we find that this subset constitutes nearly $30\%$ of the patient population. Our approach provides a natural way of uncovering this heterogeneity and thus enabling effective human-AI collaboration.