When Robots Become Too Real: Our Uncanny Perception of Robots in Light of Autistic Traits
Annelie Perruchoud
Psychology
4
Year of study:
Abstract
In the ever-growing world of technology, the prospect of engineering and designing ultra-realistic artificial agents is compelling - but comes as a double-edged sword. The uncanny valley effect (UVE) proposes that although affinity towards anthropomorphic agents grows proportionally to their human-likeness, this affinity tends to suddenly 'dip' into eeriness as that agent reaches close, but imperfect human realism. This phenomenon has gauged interest from both pop-culture and scientific fields including human-robots interaction, and even the field of autism. Autism is a neuro-developmental disorder marked by atypical social, stereotyped and sensi-motor behaviours, affecting their integration in society. Treatments under investigation demonstrate the potential of therapeutic robots; yielding significantly higher performance in developing autistic children's social skills, compared to human interventions. Remarkably, children with autism do not experience the characteristic uncanny valley 'dip' and show increased interest in robots relative to humans.
Could this difference in the UVE be extended to autism traits in a non-clinical adult population? Exploring the social cognitive and social motivational reasons why this might be the case, this exploratory research investigates the interacting effect of autistic traits in the general population on the UVE. The benefits of this approach are twofold. First, it expands our understanding of autism, its lesser variants, and genetic underpinnings, with practical applications for robot therapy in autism. Then, exploring the role of individual differences on the UVE, it informs the social, cognitive and biological roots of the UVE, while guiding the design of social robots and expanding our understanding of human nature.