It means this will kill people but fewer people than humans would and they have actual data that backs this assertion up.
The benchmark is not an alert, cautious, and sober driver because that's often not how actual people drive. So right now it's often safer to drive other times it really is safer to use autopilot, net result fewer people die.
Autopilot should not be accessible to drivers until their driving habits have been assessed and baselined. If autopilot is known to drive more safely than the individual human driver in environments X, Y, and Z then it should be made available in those environments, if not encouraged. That might not make an easy sell since a major value prop is inaccessible to many of the people who really want to use it, but it's the most reasonable, safest path.
I also imagine that cars will learn from both human and each others' driving patterns over time, which (under effective guidance) should enable an enormous improvement in a relatively short period of time.
I can grab the data, but the surprising thing about fatal crashes are the typical circumstance. It's an experienced driver driving during the day, in good weather, on a highway, in a sedan style car, sober. There are two ways to interpret this data. The first is we assume that crashes are semi-randomly distributed, so since this is probably the most typical driving condition it therefore naturally follows that that's where we'd expect to see the most fatalities.
However, I take the other interpretation. I don't think crashes are randomly distributed. That that scenario is absolutely perfect is really the problem because of humans. All a crash takes is a second of attention lapse at a really bad time. And in such perfect circumstances we get bored and take road safety for granted as opposed to driving at night or navigating around some tricky curves. And that's a perfect scenario to end up getting yourself killed in. And more importantly here, that's also the absolutely perfect scenario for self driving vehicles who will drive under such conditions far better than humans simply because it's extremely trivial, but they will never suffer for boredome or lack of attention.
It means this will kill people but fewer people than humans would and they have actual data that backs this assertion up.
The benchmark is not an alert, cautious, and sober driver because that's often not how actual people drive. So right now it's often safer to drive other times it really is safer to use autopilot, net result fewer people die.