Deathbots: Remnant Persons or Animated Personae?

Patrick Stokes (Deakin University )

Abstract: Long predicted, AI chatbots based on the ‘digital remains’ of deceased internet users have now become a reality. This technology – variously referred to as ‘deathbots,’ ‘ghostbots,’ ‘thanabots’ or ‘Interactive Personality Constructs of the Dead (IPCDs) – raises a host of ontological issues. What is the relation of a deathbot to the living user on which it is modelled? Is it an extension of personhood or agency, or a mere copy or replica? Jurgis Karpus and Anna Strasser have recently argued that a deathbot or other kind of ‘digital zombie’ may preserve a ‘sliver’ or ‘tad’ of personhood, aligning deathbots with other claimed forms of ‘remnant’ personhood. On the other hand, building on the work of Watsuji, Masahiro Morioka has recently argued for a category he calls the ‘animated persona,’ a “silent voice” appearing on surfaces such as a corpse, a brain dead patient, or even a mannequin, robot, or an object associated with a dead person. So which of these categories might best apply to deathbots? In this paper I argue that deathbots might be conceptualised as comprising both the remnant personhood of ‘digital remains’ and an animated persona generated by the affordances of AI – and that this combination creates a risk of our improperly seeing deathbots as full persons.

Bio: Patrick Stokes is Associate Professor of Philosophy at Deakin University. He is Chief Investigator (with Adam Buben, University of Leiden) for the ARC Discovery Project "Digital Death and Immortality", and the author of Digital Souls: A Philosophy of Online Death (Bloomsbury, 2021).

Zoom link available on request to Sean Bowden ([email protected])