The Credibility Problem in Human Robot Interaction

Résumé En

Commitments play a central role in human joint actions. In particular, they facilitate coordination and cooperation by stabilizing expectations, reducing the uncertainty of the interaction, providing reasons to cooperate or improving action coordination. However, commitments can only serve these functions if they are credible in the first place. In other words, commitments can only play a function in joint action as far as the participants, more often than not, comply with their commitments. This so-called credibility problem of commitments has an important counter-part in Human-Robot interaction since some of the pro-social motivations that humans exhibit in their interactions seem to be absent when they interact with robots (Sahaï et al 2017, 2019). The aim of this paper is twofold. First, we introduce the human-robot credibility problem and show how it can undermine the interaction between human and robots. In particular, we argue that the problem is especially challenging when considering how commitments are maintained and repaired during joint action. Second, we review some recent literature in psychology and philosophy of mind in order to draw different strategies that can be used in social robotics for overcoming the problem and compensate for the absence of pro-social motivation in HRI.