The author lost me with this: This is straight out of Manipulating Statistics 101. Tesla Autopilot is safer, on average, than all driving. But if you want safe, ditch the Tesla and just move to Massachusetts or Minnesota or Vermont or another state with lower fatality rates. Or maybe it isn't so simple. Maybe lower income tends to correlate with higher death rates, and maybe the purchase of a Model S tends to correlate with higher incomes. Autopilot, and all the others under development or partially rolled out by a range of manufacturers (I lust after adaptive cruise control) are impressive additions to cars, but this writer is an idiot.0.76 fatalities per 100 million miles driven
It's understandable he lost you there, but the very next sentence does say this: He is undermining his first few paragraphs there, but at least he isn't a complete idiot.Again, let's not lie with statistics. The conditions under which Autopilot, Tesla's semiautonomous driving feature, will operate are, by definition, some of the safest and most predictable situations possible. [...] Implying that Tesla Autopilot is safer than the average driver is probably like suggesting that most convicted felons are Democrats; the raw data might support the assertion, but there are other, more complex factors in play.