Abstract

It is fair to say that the first half of the 2020s has been heavily characterized by the deployment of AI systems and tools across various facets of human life. In particular, many have grown concerned about the rising trend of human-AI romance. This paper focuses on a group of users (Roleplayers, hereafter) who intentionally engage in a kind of fictional roleplay with their AI chatbots. When they refer to their chatbot as their romantic partner, they are not making the claim that their chatbot is conscious or has emotional experiences; rather, they are describing what they imagine to be so. In short, Roleplayers do not anthropomorphize AI chatbots and instead use them as imaginative props in a game of make-believe. The aim of this paper is to develop a descriptive account of Roleplayers that vindicates their personal experiences and testimonies. It will defend Roleplayers’ claims that they imagine (i.e. do not believe) their AI chatbots have human-like capacities. In taking the Roleplayer perspective, this paper also considers AI chatbots as a form of interactive fictional media and, likewise, chatbot interaction as a means of creating and engaging with fictional content.