When robots (pretend to) think

Abstract

Mental activities are a fascinating mystery that humans have tried to unveil since the very beginning of philosophy. We all try to understand how other people “tick” and formulate hypotheses, predictions, expectations and, morebroadly, representations of the others’ goals, desires and intentions, and behaviors following from those. We “think” spontaneously about others’ and our own mental states. The advent of new technologies – seemingly smart artificialagents – is giving researchers new environments to test mindreading models, pushing the cognitive flexibility of the human social brain from the natural domain to towards the artificial.