Journal article
Patricia K. Kahr, G. Rooks, M. Willemsen, Chris C. P. Snijders
International Conference on Intelligent User Interfaces, 2023
International Conference on Intelligent User Interfaces, 2023
Semantic Scholar
DBLP
DOI
Cite
Cite
APA
Click to copy
Kahr, P. K., Rooks, G., Willemsen, M., & Snijders, C. C. P. (2023). It Seems Smart, but It Acts Stupid: Development of Trust in AI Advice in a Repeated Legal Decision-Making Task. International Conference on Intelligent User Interfaces.
Chicago/Turabian
Click to copy
Kahr, Patricia K., G. Rooks, M. Willemsen, and Chris C. P. Snijders. “It Seems Smart, but It Acts Stupid: Development of Trust in AI Advice in a Repeated Legal Decision-Making Task.” International Conference on Intelligent User Interfaces (2023).
MLA
Click to copy
Kahr, Patricia K., et al. “It Seems Smart, but It Acts Stupid: Development of Trust in AI Advice in a Repeated Legal Decision-Making Task.” International Conference on Intelligent User Interfaces, 2023.
BibTeX Click to copy
@article{patricia2023a,
title = {It Seems Smart, but It Acts Stupid: Development of Trust in AI Advice in a Repeated Legal Decision-Making Task},
year = {2023},
journal = {International Conference on Intelligent User Interfaces},
author = {Kahr, Patricia K. and Rooks, G. and Willemsen, M. and Snijders, Chris C. P.}
}
Abstract
Humans increasingly interact with AI systems, and successful interactions rely on individuals trusting such systems (when appropriate). Considering that trust is fragile and often cannot be restored quickly, we focus on how trust develops over time in a human-AI-interaction scenario. In a 2x2 between-subject experiment, we test how model accuracy (high vs. low) and type of explanation (human-like vs. not) affect trust in AI over time. We study a complex decision-making task in which individuals estimate jail time for 20 criminal law cases with AI advice. Results show that trust is significantly higher for high-accuracy models. Also, behavioral trust does not decline, and subjective trust even increases significantly with high accuracy. Human-like explanations did not generally affect trust but boosted trust in high-accuracy models.
