AI 'godfather' says OpenAI's new model may be able to deceive and needs 'much stronger safety tests'

AI 'godfather' says OpenAI's new model may be able to deceive and needs 'much stronger safety tests'

Alex Wong / Staff / Getty Images

OpenAI's new o1 model is better at scheming — and that makes the "godfather" of AI nervous.

No comments

Read more