Part of me wonders if recent problems with AI (Huang leaks, OpenAI models hacking Huggingface), haven't been arranged by OpenAI in order to poison Anthropic's IPO.