How Alarmed Should We Be About AI?

I want to get into the question of how alarmed we should be your answer is maybe along some dimensions at least not as alarmed as some people are trying to get us. And I don't think that that conclusion is really derailed by the things you do because there's plenty of room for mischief and trouble, even if the AI is not going to rise up and, and destroy us the way it might in a movie or something. We should still be alarmed enough to try to have international conversations about what we're doing here and not let this be propelled by some kind of mindless arms race. The key step in the singularity is when the AI becomes able to improve itself. Once it does that