Social RoboticsRobots struggle to win back trust after repeated mistakes, Michigan study finds
Apologies, explanations and promises all failed to fully repair trust by the third error.
Researchers at the University of Michigan School of Information had 240 participants work with a robot co-worker that repeatedly let them down and then tried one of four repair tactics: apologising, denying, explaining or promising to do better. After three violations none of the strategies restored trust in full. The authors concluded that robots should be competent at a task before relying on social repair, because failed attempts can cause lasting damage. The work matters for any robot that must keep a working relationship with people over time.
Source: Tech Xplore
This archive entry summarises reporting published elsewhere; follow the source link for the full story. The summary was drafted with AI from that reporting.