OpenAI has shared some of the internal rules it uses to guide the output of its chatbot ChatGPT. This comes at a time when right-wing commentators claim the company and its AI have “gone woke.”