Title
Robotics Affordance Mixtures and the ‘Kickable’ Challenge to Anthropomorphism in Early Robot Ethics
Abstract
Although robots are often portrayed in the media as highly efficient and ubiquitous, the reality is far from this image. This is particularly true for non-industrial social robots. Recent multimodal AI systems seem to promise new advances in robotics that might begin to change this and bring robots into more and more environments. However, the normative conceptual space for how we evaluate the humans’ treatment of robots is not well prepared for such an expansion. Most discussions are still motivated by graphic scenarios that depict humans mistreating robots (e.g. kicking a robot) and by anthropomorphism to describe human experience in HRI. As such, they result in concerns that the mistreatment of robots could have undesirable consequences or implications for human moral practices. Thus, they normatively problematize mistreatments. These approaches fail to consider broader, everyday interactions. These include the use of robots in everyday life, where robots come with unique affordance mixtures, don’t necessarily resemble/represent other beings, and are perceived directly (rather than fictionalized or anthropomorphized). As such, robots can be treated differently from people/animals. I introduce “robotic affordance mixtures” to describe the versatile capacities of robots. And I show how current approaches fail to account for the novel and versatile affordance mixtures that robots can offer; the novel and dynamic ways in which human-robot interactions can occur; the use of robots in many ways as products, thanks to affordance mixtures, while robots participate in the social realm. Then I apply ecological psychology and the conceptual tool of ‘sociomorphing’ to emphasize how situation, context, dynamics, interaction and direct affordance perception matter in HRI, where robots play an active role rather than being passively anthropomorphized/fictionalized. With such tools, normative discussion can begin to distinguish between an affordance treatment of a robot from a mistreatment, an everyday use from an abuse. So, there is a need for more cognitive-scientific and conceptual insight into what I refer to as ’early robot ethics’. In particular, into the indirect robot moral patiency debate, which has been stuck on graphic scenarios of mistreatment, and stuck on limited descriptive concepts like anthropomorphism. But now the debate needs to move away from such limited and inaccurate HRI formulations based on anthropomorphism—where interaction is understood passively and morally loaded descriptions of actions towards robots are taken for granted (‘kicking a robot is mistreatment’). We need to challenge these assumptions and move towards more nuanced investigations of HRI by focusing on mundane everyday interactions—rather than graphic scenarios—with novel and versatile robotic affordance mixtures.
About Arzu
Arzu Formánek is a researcher at the Fraunhofer IPA-Institute for Manufacturing Engineering and Automation, a researcher at University of Stuttgart, and a PhD candidate at University of Vienna (Supervisors: Mark Coeckelbergh and Mark Bickhard. External Advisor: Sven Nyholm).