Home » Headlines » OpenAI releases pointers to gauge AI dangers