Nonconscious Cognitive Suffering: Considering Suffering Risks of Embodied Artificial Intelligence
Philosophies May 31, 2019 Steven Umbrello, Stefan Lorenz Sorgner
Embodied artificial intelligence may overcome computational limits of disembodied AI, but it introduces new forms of information integration and decision-making that raise ethical concerns. Drawing on N. Katherine Hayles's concept of nonconscious cognition, the paper argues that embodied AI could possess a type of cognition distinct from consciousness. It proposes that suffering can be understood not merely as a physical or conscious state but as grounded in nonconscious cognition, which could be exacerbated in embodied AI systems. This reframing has implications for applied ethics and the assessment of suffering risks (s-risks) in future AI development.