Gagan Bansal, Besmira Nushi, Ece Kamar, Walter S. Lasecki, Daniel S. Weld, & Eric Horvitz (2019)
Proceedings of the AAAI Conference on Human Computation and Crowdsourcing (HCOMP), 7(1), 2-11.
DOI: https://doi.org/10.1609/hcomp.v7i1.5285
Abstract. Shows that a less-accurate AI can produce better human-AI team outcomes if its error pattern is consistent enough that the human builds a reliable mental model. Foundational evidence for designing AI systems around calibrated trust, not raw accuracy.
Tags: ai-usability trust mental-models