If I rate my own AI predictions, every wrong one was on the optimistic side (e.g. in 2018 my timeline was mass-deployment of self-driving was 2023, which sounded extremely pessimistic back then). And people think I'm a skeptic (which I am not — historically I've been slightly