You're forgetting that they are also training with real data from the 100+ million miles they've driven on real roads with riders, and using that data to train

You're forgetting that they are also training with real data from the 100+ million miles they've driven on real roads with riders, and using that data to train the world model AI. > there's probably no examples in the training data where the car is behind a stopped car, and the driver pulls over to another lane and another car comes from behind and crashes into the driver because it didn't check its blindspot This specific scenario is in the examples: https://videos.ctfassets.net/7ijaobx36mtm/3wK6IWWc8UmhFNUSyy4kCy/4595f848baa762961c945a8e080bbc16/fast_long_rollout_1.webm https://videos.ctfassets.net/7ijaobx36mtm/3wK6IWWc8UmhFNUSyy... It doesn't show the failure mode, it demonstrates the successful crash avoidance.