Robots as Social Actors: Aurora and the Case of Autism

This paper discusses the role of predictability and control in robot-human interaction. This involves the central question whether humans are good models for synthetic (social) agents. Design issues based on cognitive accounts towards robot-human interaction are discussed with respect to the author’s recent work on building interactive robotic systems as remedial tools (teaching devices) for children with autism, a project which crucially requires a careful analysis of human-robot relationships, and which cannot rely on the ‘natural’ tendency of humans to be interested in agents that appear life-like and social, a factor which is currently exploited extensively in the believable agents research community. The paper will argue that using humans as models for creating believable technology (e.g. making the robot life-like) is not a universal solution for designing robotic social actors.