From Wikipedia
OpenAI plans to add safe completions to its next model to reduce risks of AI-generated medical guidance
