Juraj Hvorecký: AI and unconsciousness

Invitation pdf

Abstract

Advances in generative artificial intelligence (GenAI) have reignited debates about the possibility of artificial systems having consciousness. Our contribution will shift the perspective of current views on this issue. We begin with the genesis of the current debates on consciousness in GenAI networks and the basic arguments in its defence. We then present counterarguments using actual case studies from several domains of demonstrated failures of GenAI that support scepticism about their consciousness. In the final section, we introduce a new approach that rejects the attribution of consciousness to GenAI systems, yet considers them an interesting model of unconsciousness. We will show how multimodal experiments with these systems support the idea of an analogy with the processing of unconsciousness in humans, and we will also point out the ethical and social implications of such an analogy.