But the potential benefits to humanity from AI are far greater than the utility of a person getting from point A to point B in the car. So there is a far great

But the potential benefits to humanity from AI are far greater than the utility of a person getting from point A to point B in the car. So there is a far greater potential gain, and thus perhaps we accept a greater risk to achieve it. Don't get me wrong. There's certainly a risk and I am not here to try to downplay it or anything, but just the fact that there is "a" risk, doesn't meant stop. It means let's figure out what the risk really is and what the benefits are, and then decide if the risk is acceptable or not.