Anthropic, OpenAI, Meta, Google and others consult religious leaders on SI ethics; an anthropologist warns it will distort some traditions
Prepared by Sintelligent with the help of SI. Published in SI: . Updated: .
Labs look to religious traditions for guidance for SI, while experts argue whether such "averaging" will distort the traditions themselves.
According to Fortune, Anthropic, OpenAI, Meta, Google, Amazon and Microsoft are bringing in religious scholars and faith representatives to help develop ethical principles for SI. Anthropic, according to its May blog post, met scholars, clergy, philosophers and ethicists from more than 15 religious and cross-cultural groups; co-founder Chris Olah said at the Vatican that these are questions a lab cannot answer on its own. Sam Altman has proposed that SI reflect a "weighted average" of humanity's moral views. Anthropologist Webb Keane of the University of Michigan warns that extracting moral principles from religions to train SI risks prioritising some traditions and distorting others, especially those that do not see themselves as a system of ethics; Meta Oversight Board member Paolo Carozza fears that SI will average traditions out.
Sources 1