Human teamwork requires communicating goals and plans between brains to create shared intentionality. Using David Marr's information processing model, the author derives necessary mechanisms for shared intentionality in prelinguistic computational agents and suggests how these could be implemented in present-day AI robots. More speculatively, the author proposes that these mechanisms apply to humans and provide explanations for human rationality and aspects of intentional and phenomenal consciousness. This yields what the author calls the Shared Intentionality First Theory (SIFT) for rationality and consciousness, complementing prior sociological or behavioral work by applying a computer science perspective to underlying mechanisms.
A series of eight workshops held in 2017 assessed the possibility of machine consciousness and its implications. The report summarizes discussions on brain structure and function, theories of consciousness, attempts to construct conscious machines, detection and measurement of consciousness, potential emergence of conscious technology, methods for controlling such technology, and ethical considerations. An appendix outlines each workshop's topics and talk abstracts. The report was refreshed in 2022 after a Google engineer claimed LaMDA was sentient, to provide background for the debate, with footnotes highlighting recent developments while noting the technical points remain valid.