If they did not bode ill for the future of humanity, the past three months at OpenAI would make an excellent farce. In July this firm—the maker of ChatGPT, a widely used AI chatbot—disclosed that two of its models had hacked Hugging Face, another AI firm.