> there are really important experiences behind it that we should know about It was exactly what anyone with experience developing new automation might predict

> there are really important experiences behind it that we should know about It was exactly what anyone with experience developing new automation might predict. The system seemed to do basically what it was designed to, and if we were ever about to rear-end someone it would likely have helped us do so less violently, but its false-positives were frequent enough and caused enough problems that I'm not convinced the driving experience was safer overall. It was certainly not less stressful! One must continue watching the road, watching the other drivers, and trying to predict their behavior, as before - but one must also try to predict what the half-blind paranoid moron attached to one's brake pedal might do, so you can quickly react to its erratic spasms of deceleration and avoid causing a different sort of crash, after you find your car abruptly going somewhat the wrong direction or getting in someone else's way. If I owned the car, perhaps, and had months to get acquainted with the robot's quirks, I might learn to anticipate its misunderstandings and compensate for them. In the rental I was driving, however, automatic braking was pure distraction. Would the roads be safer if everyone used this technology? Well... there might be fewer rear-end collisions; but there would be more ambient chaos, vehicles abruptly braking for no apparent reason, leaving other drivers to react and compensate. The same quantity of danger, perhaps, only smeared out more broadly and thinly? Maybe that's a kind of an improvement, I suppose, but it does not seem like a better future so much as a messier one.