This seems like playing with semantics. You could say safety isn't a priority, or that it's the highest priority, depending on what you define as "safety". It'

This seems like playing with semantics. You could say safety isn't a priority, or that it's the highest priority, depending on what you define as "safety". It's actually interesting that discussions about why people don't feel safe when they're not the driver are usually not very coherent. Fear of flying is the classic one: you have zero control, including the lack of ability to end the experience. Somehow, despite a century of commercial flight, aviophobia / aerophobia is still poorly understood and hasn't been studied very thoroughly. As a result, some fairly ludicrous explanations get tossed around, for instance that the fear is all due to claustrophobia (even in people who aren't otherwise claustrophobic) or lack of understanding of physics (barely anyone really understands physics, but no physics-illiterate person gets scared by, say, walking down a flight of stairs). Maybe in your case you really are distrustful because of an entirely rational understanding of how software is made and what edge cases are. But like physics, very few normal people have any understanding of building software. And like flying, the real reason that the vast majority of people aren't -- and probably won't be, for a long time -- comfortable with self-driving cars is simply, lack of control. "Safety" is not really the relevant problem here. Lack of control is. One horrifying, highly publicized accident a year could completely put off the entire population from self-driving, nevermind if the safety of self-driving is two, ten or a thousand times better than human controlled.