That’s fine, but even in your post I’m seeing some easy-to-claim stuff that has no solid basis. Are you sure that the programmers cannot explain why a chatbot errors out? Really? Also, who said anything about the emotional state of an AI? That’s hardly even possible because it doesn’t have an endocrine system. We may have strong emotions the way we do to help with memory formation and retrieval as much as anything else. That’s not a problem for a machine. What’s a plausible way we get destroyed? Does AI own the corporations too? How do I lose power, internet, food, etc,? The nuclear terminator version seems impossible unless we are going talk about hacking brains and adjusting behavior like crazy people think is possible.