Ethics of Social Robots as Seen by Their Designers and Users

• In Event: Constellation Thinking and AI Ethics

Abstract

A large number of designers and users share the same horizon: The possibility of establishing a personal and lasting relationship with social robots. Yet the gap between this horizon and the current reality remains large. Are some ethical concerns expressed in the public space too out of sync?
Several actors consider that certain ethical questions (such as attachment) are not relevant in view of the still very limited interactional skills of social robots (regarding facial recognition for instance). Do the current technical limitations make public debates on ethics irrelevant? Does this mean that the question of ethics does not yet have to be asked?
This observation rather invites us to focus on ethics in practice, in the ordinary deployment of social robotics. Danah Boyd and Madeleine Elish (2018) point out that the mismatch between the imagined capacity of a technology and its current reality often leads public conversations about ethics to focus on “hypothetical extremes” instead of the “current ethical dilemmas that need to be faced here and now”.
To what extent does the sociological study of the “here and now” of social robots make it possible to identify the real ethical issues? We propose to compare the ethical concerns regarding social robots as they are formalized in the gray literature and as they emerge in situated experiments.
This analysis is based on an empirical survey on the design and uses of social robots for children with ASD (autism spectrum disorder) in France. A comparative ethnographic study (interviews, observation and documentary analysis) was conducted on two contrasting social robots (LEKA and NAO).
It shows that the study of ethics from the point of view of designers and users of social robots allows to shift, but also to refine, the terms of the public controversy on different issues (anthropomorphism and confusion between human and non-human, relations and attachment, substitution and dehumanization, decision-making and accountability).

Author

• Céline Borelle, Orange Labs