In that video, it looks like the driver had about 1 second to react to the Tesla's mistake, to avoid hitting the barrier.
Dumb question: if I have to constantly stay vigilant with 1-second reaction time because at any time my assisted-driving car might try to kill me (exactly how changes every time the software updates) -- isn't it less effort to just drive my car myself, so I mostly just have to worry about the drivers around me?
Not dumb at all, it is the big unspoken elephant in the room. It is clear that the current, and next iteritave version of autopilot, self driving, etc. are woefully behild the hype. While in theory a "perfect" ai driver is better, is it true in the chaos of real world traffic? Will drivers even want to be subject to NTSB level scrutiny when an accident happens? There are more questions than answers still.
Dumb question: if I have to constantly stay vigilant with 1-second reaction time because at any time my assisted-driving car might try to kill me (exactly how changes every time the software updates) -- isn't it less effort to just drive my car myself, so I mostly just have to worry about the drivers around me?