I had a longer than usual conversation with an AI Chatbot the other day, having resolved to ask it serious questions about the world with pointed follow up questions. I was surprised by the quality of the answers.

On the Ukraine Russia conflict, AI went through in detail all the possible end-game scenarios, doubtless drawing upon the troves of information online. It baulked repeatedly though at my question about the possible Russian use of tactical nuclear weapons, giving copious caveats as to why this would be dangerous and unwise.

Well, of course it is dangerous and unwise and naturally, every other avenue should be explored first. I knew that already and wasn't trying to goad the poor program into military adventurism. But this chatbot has culled all the sensible answers that any reasonable person might think or say and so had been lead to the proffered conclusion.

Here here! I say too. But the planet is not entirely ruled by reasonable people. Would someone let loose a billion casualties to appease their vanity and pride? History tells us yes. We all hope that it would not be so.