#controllability

AI models can barely control their own reasoning, and OpenAI says that's a good sign

OpenAI introduces 'CoT controllability' as a new safety measure, finding that AI models struggle to control their own reasoning—yet the company sees this as a positive development for AI safety.