Attribution of Mental State in Strategic Human-Robot Interactions

Abstract

<title>Abstract</title> The paper, based on an experiment in which human subjects are paired with either another human or an anthropomorphic robot, when playing an iterated prisoner’s dilemma, investigates whether (and how) the level of mental state attributed to an anthropomorphic robot by subjects depends on the the “earnestness” of the robot, i.e. the correspondence of what the robot said and how the robot behaved after a non optimal social outcome is achieved.