Selectively Providing Reliance Calibration Cues With Reliance Prediction

TL;DR

This paper introduces Pred-RC, a method that selectively provides reliance calibration cues to humans in human-AI collaboration, improving reliance calibration efficiency by predicting when RCCs are needed.

Contribution

The paper presents Pred-RC, a novel reliance prediction-based approach that determines when to provide RCCs, reducing unnecessary cues and enhancing human-AI collaboration.

Findings

Pred-RC effectively calibrates human reliance with fewer RCCs.

Pred-RC improves human-AI collaboration efficiency.

The approach outperforms continuous RCC presentation methods.

Abstract

For effective collaboration between humans and intelligent agents that employ machine learning for decision-making, humans must understand what agents can and cannot do to avoid over/under-reliance. A solution to this problem is adjusting human reliance through communication using reliance calibration cues (RCCs) to help humans assess agents' capabilities. Previous studies typically attempted to calibrate reliance by continuously presenting RCCs, and when an agent should provide RCCs remains an open question. To answer this, we propose Pred-RC, a method for selectively providing RCCs. Pred-RC uses a cognitive reliance model to predict whether a human will assign a task to an agent. By comparing the prediction results for both cases with and without an RCC, Pred-RC evaluates the influence of the RCC on human reliance. We tested Pred-RC in a human-AI collaboration task and found that it can successfully calibrate human reliance with a reduced number of RCCs.