Good points. You can't make determinations from news flashes. And it would be interesting to see solid numbers/statistics from the NTSB. Your likely correct.And, since the thread has been revived: I was simply referring to the frequency of accidents, and therefore eliminating cherry picking. But you could easily normalize the data using vehicle miles traveled. I'd be willing to bet the "self-driving" cars have a much lower accident rate, despite the "freak" accidents attributable to malfunctions.
On a different note, it's billed as an assist. The driver is burdened to always maintain control. And that's the fallacy of the current state of the technology.
This guy was on his phone. I recall an earlier accident where the driver was watching DVDs. What are you supposed to do - watch the autopilot drive? Accidents can be measured in tenths of seconds. If you not actively driving yourself, your never in control. Your at the mercy of the technology.