> I don't think there is anything to indicate that Tesla's approach is leading to more deaths.
The safety claims Tesla/Musk make are debatable. There is a lawsuit over getting the raw data NHTSA used to make the "autosteer reduces collisions by 40%" [1]. AEB also reduces the collision rate by 40 percent, which is mentioned in the same NHTSA report [2]. And, supposedly AEB and autosteer were activated around the same time, so it might not be easy to tell which provided a benefit.
> I would bet that autopilot is still well in the positives in "lives saved" versus "lives ended"
You need a lot of data to show that. 1 death per 86 million miles driven is the average in the US, and that's across all types of vehicles, new and old, including motorcycles, bad weather (where autopilot is not likely to be engaged), and on all kinds of roads (autopilot is mostly engaged on divided highways).
There was also a death in China that looks like autopilot [3].
The safety claims Tesla/Musk make are debatable. There is a lawsuit over getting the raw data NHTSA used to make the "autosteer reduces collisions by 40%" [1]. AEB also reduces the collision rate by 40 percent, which is mentioned in the same NHTSA report [2]. And, supposedly AEB and autosteer were activated around the same time, so it might not be easy to tell which provided a benefit.
> I would bet that autopilot is still well in the positives in "lives saved" versus "lives ended"
You need a lot of data to show that. 1 death per 86 million miles driven is the average in the US, and that's across all types of vehicles, new and old, including motorcycles, bad weather (where autopilot is not likely to be engaged), and on all kinds of roads (autopilot is mostly engaged on divided highways).
There was also a death in China that looks like autopilot [3].
[1] http://www.safetyresearch.net/blog/articles/quality-control-...
[2] https://static.nhtsa.gov/odi/inv/2016/INCLA-PE16007-7876.PDF
[3] https://www.youtube.com/watch?v=fc0yYJ8-Dyo