I don’t understand how can anybody take it seriously an approach that relies on statistics for self driving cars. Even if it is 95% correct who would risk getting killed 5% of the time while driving?
Of course not 95/5%, but we rely on statistics that all kinds of bad cases will likely not happen every time we enter traffic. What the right statistic is, how to get good data on it, and mentally accept it and the loss of perceived control, that's going to be interesting.
