I agree that this is mostly OpenAI’s fault, though I also think people posting strong claims about LLMs online have a responsibility to know slightly more than the average user.