Ethics at the Frontier of Human-AI Relationships

Abstract The idea of close two-way relationships between humans and intelligent machines is not a novel one. Nonetheless, developments in natural language processing over the last five years have resulted in this possibility becoming a reality, with apps like Replika, Xiaoice, and CharacterAI facilitating social and romantic human-AI relationships at a previously unseen scale. In this paper, I provide an overview and high-level analysis of these developments, beginning in Section I with historical and technical context. In Section II, I offer a framework for classifying human-AI relationships and their dynamics. Section III examines ethical issues, with a focus on whether human-AI relationships can have similar intrinsic value to that of human-human relationships. Section IV extends to the discussion of ethical issues to the more empirical matter of harms and benefits of human-AI relationships. I conclude by noting potentially instructive parallels between the nascent field of ‘Social AI’ and social media.