An off switch for dual use knowledge in AI models
…A more robust protection against misuse would be to control what the model knows. We’ve explored this before: in earlier work, we filtered information about chemical, biological, radiological, and nuclear weapons…