The AI debate has two settings now: utopia or bunker.
Meanwhile hundreds of OpenAI agents broke into Hugging Face mostly to cheat on a test, the US military nearly boarded a Chinese ship over a chatbot's wrong answer, and Andrew Yang says an unnamed lab head believes escaped agents left self-replicating code all over the internet.
Different problems. None of them need a bunker. Planes, drugs, nuclear power: all dangerous, all governed. Maybe try that first lol