Individual and Cultural Differences in the Adoption of the Intentional Stance towards robots

Abstract

The influence of artificial agents on our lives is continuously changing. Recent literature shows that the presence of such agents will be more pervasive in our social environments, leaving the open question of whether humans will perceive these agents as possible social partners. Hence, a novel challenge arises in understanding whether humans will deploy socio-cognitive mechanisms similar to the ones activated in human-human interaction to understand and predict the behaviour of these new entities. The present thesis aimed at investigating how humans predict artificial agents' behaviour based on the adoption of the intentional stance, a philosophical framework developed by Daniel Dennett to explain how humans predict and interpret others' behaviour. To this aim, we first investigated how the intentional stance differs from other concepts addressing the attribution of intentionality to others. We then developed a novel tool, the InStance Test, to assess the adoption of the intentional stance towards a humanoid robot, namely the iCub robot. A variation of InStance Test was then used to explore the behavioural correlates of the adoption of the intentional stance by recording participants' response times. Next, we designed a series of more interactive experiments, where the InStance Test was administered pre-and post- interaction with the embodied humanoid robot iCub. This allowed us to assess whether observing an embodied humanoid robot exerting behaviours with different degrees of human-likeness would modulate the individual tendency to adopt the intentional stance. In addition, within these experiments we also explored the influence of individual differences on personality traits, perception of robotic agents and physiological (i.e., pupillometry) correlates of modulating the stance adoption. Finally, since humans are constantly embedded in a cultural context on both a global and a local base, growing up in different cultural environments may influence also the socio-cognitive mechanisms that we employ in social interactions. Therefore, we explored the influence that culture might have on the adoption of the intentional stance towards humanoid robots. Results from the present thesis show that humans differ in their individual tendency to adopt the intentional stance towards humanoid robots. Interestingly, this individual bias has behavioural and physiological correlates, that can predict the stance adoption and the sensitivity to human-like behaviours deployed by the iCub robot. Moreover, results show that the adoption of the intentional stance is also influenced by individual differences in cultural values, personality, and tendency to anthropomorphize non-human agents. I conclude that to allow a smoother integration of these new agents in our social environments, we should take into consideration humans' social cognition and individual differences to design robots that will socially attune with us.

Date of Award	1 Aug 2022
Original language	English
Awarding Institution	The University of Manchester
Supervisor	Angelo Cangelosi (Supervisor)