3.2 fatalities per billion km driven by humans. So far 1 fatalities per 4 million miles by automated vehicles, which when you do the conversions... isn’t better. It’s not a lot to work with, but for the people speaking in stats, the actual stats don’t support them. A lot people in this thread seem to be cashing in a human life against future hypothetical gains which have in no way been demonstrated. I sincerely hope that I’m not the only one disturbed by that.
Not sure it's that clear. Using data from Wikipedia, I got 11 deaths per billion km. The 4 million miles are just from Weymo, not all autonomous cars in total. (6 million+ https://medium.com/self-driving-cars/miles-driven-677bda21b0...) And the confidence level on the 1 event over 6+ million miles is not great either.
Again, I’m responding to people touting the safety of autonomous vehicles based on a tiny sample. Live by the sword, die by the sword. At this point the truth is that we’re surrounded by unknowns, and yet post after post features people claiming inhuman safety records for a nascent technology.
