Overconfidence in and abuse of the Autopilot system can't be excused. But the article still makes me wonder: if many Autopilot accidents also involve "regular" risk factors like alcohol; how do the accident numbers compare when comparing drunk Autopilot drivers vs. drunk manual drivers?
It's easy to criticise any system by zooming in on specific accidents involving it. It's probably harder to assess whether a system will yield a net positive or negative impact on overall accidents.
This situation doesn't surprise me, after reading this skeptic article on self-driving in May, and especially watching this linked video from March.
Overconfidence in and abuse of the Autopilot system can't be excused. But the article still makes me wonder: if many Autopilot accidents also involve "regular" risk factors like alcohol; how do the accident numbers compare when comparing drunk Autopilot drivers vs. drunk manual drivers?
It's easy to criticise any system by zooming in on specific accidents involving it. It's probably harder to assess whether a system will yield a net positive or negative impact on overall accidents.