Human-AI Collaboration in Decision-Making: Beyond Learning to Defer

TL;DR

This paper reviews the limitations of current Human-AI collaboration frameworks like Learning to Defer, highlighting challenges in real-world deployment and proposing directions for future research to improve decision-making synergy.

Contribution

The paper critically analyzes existing HAIC methods, especially L2D, and identifies key practical limitations and future research opportunities beyond current approaches.

Findings

L2D requires predictions from humans for all instances

Current methods struggle with real-world deployment issues

Identifies future research directions in HAIC

Abstract

Human-AI collaboration (HAIC) in decision-making aims to create synergistic teaming between human decision-makers and AI systems. Learning to defer (L2D) has been presented as a promising framework to determine who among humans and AI should make which decisions in order to optimize the performance and fairness of the combined system. Nevertheless, L2D entails several often unfeasible requirements, such as the availability of predictions from humans for every instance or ground-truth labels that are independent from said humans. Furthermore, neither L2D nor alternative approaches tackle fundamental issues of deploying HAIC systems in real-world settings, such as capacity management or dealing with dynamic environments. In this paper, we aim to identify and review these and other limitations, pointing to where opportunities for future research in HAIC may lie.