Buddhism, Consciousness, and the (Im)Possibility of Ethical AI

This article explores the relevance of classical Theravāda Buddhist doctrine to the present-day development of artificial general intelligence. Specifically, it addresses the interconnected possibilities of machine consciousness and machine agency. The first section consists of a philosophical exploration of the notion of artificial consciousness in light of ordinary language considerations. This is followed by a Buddhist theoretical account of the conditions necessary for the arising of consciousness, relying in good part upon the medieval Abhidhamma commentary, the Abhidhammattha Saṅgaha. Serious doubts are raised as to whether consciousness could ever be created in a machine environment. The final section examines the possibility of machine agency in relation to Buddhist understandings of action (kamma). Here, it is argued that if conscious machines ever were developed, whatever agency they might demonstrate would be amoral in nature and reflective of the values of their corporate and state programmers. Their development would pose considerable dangers to living beings. While the main argument of the article is made in Buddhist terms, it is supported throughout by more general philosophical considerations and with reference to some of the relevant scientific literature.