It is not just the left that is in love with AI. Many on the Right and in the center are equally attracted. I can appreciate appreciating the promise of benefit, but (or maybe I should have said But as it's a big ass But) I am totally down with the reality, they can be both too smart and too dumb for our own good. We already know science fiction often becomes science fact and inherit hubris of man doth bite deep at times, and I have been a science fiction reader since being a little kid, cutting my teeth on Azimov in early grade school. Also, not to be forgotten is possible actions of AI due to bad human actors tweaking code with intent. The power of AI is presently unlimited in scope, so equally unlimited in potential danger and damage. It very well might not be as simple as flicking a switch and turning it off and no matter how fast you are, you are not faster than a speeding electron and will be too late, sooner or later. Presently, we have already seen it can lie (or more accurately) create false date or output for human consumption and this is just the infancy of the technology. It can steal (or more accurately) appropriate data to create it's sudo-intellectual output, with one company being sued, even as we speak. There simply are no rules, nor over site at the present time and certainly no Azimov inspired "3 laws", even those did not function as intended at time or control resulting damage.