Digital Avatars and the Ruist Ontology of Selfhood

Abstract Would Ruism encourage filial piety towards a robot? This question is not an absurd one. The large language models (LLM) we interact with daily are becoming increasingly Turing-indistinguishable from human interlocutors, and this trend is likely to continue. Such models could take the lifetime digital footprint of a human as training data from which to generate avatars which even close relatives would find to be increasingly Turing-indistinguishable from their biological loved one. This prospect has implications for how we go about “theorizing persons,” an enterprise recently undertaken by Roger Ames. Specifically, this paper will argue that it strengthens the hand of those who resist the strongest construals of the relational ontology of selfhood. We can easily, for instance, conceive of outwardly vibrant and creative relational networks the constituent nodes of which have nonetheless become partially or totally zombified by avatar proliferation. This underscores the importance of consciousness for the Ruist ontology of selfhood, over and above the consummate enactment of social roles. We will show that this is in fact completely in accordance with canonical Ruist concerns. Ontologically, it will be reasonable to maintain that this substrate is substantial rather than fundamentally relational, although not substantial in any Cartesian sense.