Interesting challenge for us then--if we're pouring money into self-driving cars and constantly making strides there, why can't we do the same for fully automat

Interesting challenge for us then--if we're pouring money into self-driving cars and constantly making strides there, why can't we do the same for fully automated self-running nuclear reactors. If there's a future where we trust cars 100% to drive every family safely around town, we should be able to trust reactors to run themselves and fail safe. I don't see why a human brain has to be in the critical loop of reactor operation, other than tradition or legal CYA.