Who Should I Trust: AI or Myself? Leveraging Human and AI Correctness Likelihood to Promote Appropriate Trust in AI-Assisted Decision-Making

TL;DR

This paper introduces methods to improve human trust calibration in AI-assisted decisions by considering both human and AI correctness likelihood, leading to more appropriate trust levels.

Contribution

It models human correctness likelihood and develops strategies to explicitly and implicitly calibrate human trust based on both human and AI performance.

Findings

Strategies increased appropriate human trust compared to AI confidence alone

Modeling human CL improved trust calibration effectiveness

Experiment with 293 participants validated the approach

Abstract

In AI-assisted decision-making, it is critical for human decision-makers to know when to trust AI and when to trust themselves. However, prior studies calibrated human trust only based on AI confidence indicating AI's correctness likelihood (CL) but ignored humans' CL, hindering optimal team decision-making. To mitigate this gap, we proposed to promote humans' appropriate trust based on the CL of both sides at a task-instance level. We first modeled humans' CL by approximating their decision-making models and computing their potential performance in similar instances. We demonstrated the feasibility and effectiveness of our model via two preliminary studies. Then, we proposed three CL exploitation strategies to calibrate users' trust explicitly/implicitly in the AI-assisted decision-making process. Results from a between-subjects experiment (N=293) showed that our CL exploitation strategies promoted more appropriate human trust in AI, compared with only using AI confidence. We further provided practical implications for more human-compatible AI-assisted decision-making.