AI as epistemic media and the future of subjectivity

Abstract

For scholars of the social sciences and humanities, the most pressing question concerning AI is its ramifications for human beings: what would it mean to be human in a world increasingly populated by AI? This, I suggest, requires us to divert from the ontological question of “what is AI?” and seek a more human-centric mode of inquiry. In this essay I would like to lay out one such possible route for inquiry. I begin by identifying the shortcomings of the common historiography of AI, which narrates a progressive development of an autonomous thinking machine from its roots in mythology up to innovations in digital technology. Instead of such history, which harks back hardly a century and culminates in the development of deep learning, I suggest thinking of AI in the context of a much longer, non-digital history of epistemic media. Epistemic media extend the cognitive abilities of human users. They facilitate the creation of new forms of knowledge about the world and about the self. For centuries now, epistemic media have offered devices for creating subjectivity by allowing reflexivity, which grew out of a dialogue between the machines and its human users. But they were not autonomous thinking machines. By thinking about AI as part of that family of media devices we can start to outline a few key axes along which questions concerning the ramifications of AI to what it means to be human can be posed.