References & Citations
Computer Science > Artificial Intelligence
Title: Do Humans Trust Advice More if it Comes from AI? An Analysis of Human-AI Interactions
(Submitted on 14 Jul 2021 (v1), last revised 1 Jun 2022 (this version, v2))
Abstract: In decision support applications of AI, the AI algorithm's output is framed as a suggestion to a human user. The user may ignore this advice or take it into consideration to modify their decision. With the increasing prevalence of such human-AI interactions, it is important to understand how users react to AI advice. In this paper, we recruited over 1100 crowdworkers to characterize how humans use AI suggestions relative to equivalent suggestions from a group of peer humans across several experimental settings. We find that participants' beliefs about how human versus AI performance on a given task affects whether they heed the advice. When participants do heed the advice, they use it similarly for human and AI suggestions. Based on these results, we propose a two-stage, "activation-integration" model for human behavior and use it to characterize the factors that affect human-AI interactions.
Submission history
From: Kailas Vodrahalli [view email][v1] Wed, 14 Jul 2021 21:33:14 GMT (2279kb,D)
[v2] Wed, 1 Jun 2022 22:26:39 GMT (2691kb,D)
Link back to: arXiv, form interface, contact.