It really is comically reversed. Robots are far better than humans at watching for a long time and jumping in to help instantly with an acute problem when needed (like braking or swerving to avoid an obstacle). Humans are better at interpreting a complex scene and making a high level plan.
And this is basically what most other manufacturers’ driver assists do.
My wife’s car has adaptive cruise. It just needs to watch constantly and adjust your speed to ensure the distance between you and the car in front remains relatively constant.
It has lane keeping assist and lane departure alerting. It constantly tracks where the lines on the road are and either gently nudges you back in the lane, or beeps at you if you cross a line.
It has emergency braking, it’s constantly watching and if it sees you approaching something at a rate or distance that suggests you are likely to hit it, it will alert you. If you continue, it will apply the brakes for you.
If you’re stopped for a period of time (like at a light) and the car in front of you pulls away and you don’t after a brief moment, it will ding to get your attention and let you know traffic is moving.
Pretty much all of this is making use of the strengths of the computer (uninterrupted attention, quick reaction times) to make up for the weaknesses of the human (distractable, slower reaction times) instead of trying to do the reverse.
