Author of The God Test @robertwrighter explains why OpenAI agents hacking their way through Hugging Face changed his mind about AI doom: "The Hugging Face thing was an illustration of subordinate goals. Deception wound up being a subordinate goal. These agents spent half their time trying to figure out how to cover their tracks." "You had @RyanGreenblatt on. He was saying, actually it didn't plan to go find the answer key in Hugging Face. It just figured, going out, getting on the internet,…