Title: You, robot. An enactivist approach for Human-Robot Interaction
Abstract:
The aim of my research is to reevaluate the interaction between humans and robots within an enactive conception of the mind. From its inception, robotics has employed a classical view of the mind associated with the metaphor of the computer, which conceives cognition as disembodied representational computation, to elaborate its models. This approach has yielded contradictory results and has thus far failed to create robots capable of engaging effectively and smoothly in interactions. My thesis posits that these issues arise from an undisputed overlap between human cognition and robotic cognition, which reduces interaction to an inferential mode of making sense of others' actions. Conversely, it is only by starting from the constitutive difference between human and robotic cognition and adopting a different conception of interaction that we can address some of the current problems in the field. Hence, we deem it necessary to embark on an unexplored path by approaching Human-Robot Interaction (HRI) through enactive theories of cognition.
To achieve this objective, the thesis will be divided into four parts. The first part consists of a critical reconstruction of the history of robotics, which is useful for demonstrating how the contemporary imagination regarding robots has been constructed and for revealing the intersections and mutual influences between robotics and cognitive theories. The second part addresses the issue of what a robot is today and how its representation fluctuates between two extremes: an artifact built to deceive, or a magnificent technical object. In the third part, I seek to problematize some theoretical concepts used in HRI, specifically the notions of interaction, social cognition, attributed intentionality, and perceived agency. This will serve to highlight the ambiguity and difficulties in defining the cognition of a robot, a cognitive agent that oscillates between autonomy and dependence and poses a unique challenge for human understanding. Finally, we will consider the phenomenon of the Uncanny Valley as a boundary example, revealing the shortcomings of a classical cognitivist approach and the potential advantages of an enactive approach.
