Apollo: the four claims of a scheming safety argument
AI summary
To develop frontier AI safely, developers must be able to demonstrate that their models are not scheming, i.e., not covertly opposing them while pursuing unintended goals. New post: the four claims any scheming safety argument must make, and the resources embedded evaluators need to verify these claims 🧵
Full text · Original
To develop frontier AI safely, developers must be able to show that their models are not scheming, i.e. covertly working against them in pursuit of unintended goals.
New post: four claims any scheming safety case must make, and the resources embedded evaluators need to verify them 🧵
