I think that's a little bit of a weird way to look at the probabilities. Sure, for a one-off activity I might look at 1 in 7,700,000 and decide that's an acceptable risk. But many people in the US take several car rides per day.

At, say, 4 rides per day, that's about a 1 in 5300 chance of death over a single year. That's still small, but not that small. Someone in a decent-sized town or city could expect to lose someone they know once every few years with those odds.

We know what the rate of deaths are: 1 in 8000; roughly 40,000 over 320,000,000.

Slightly less than the rate of suicide; and slightly more than half the number of fentanyl deaths. And a smaller fraction of medical mistake deaths. (Of course, none of the risk is evenly distributed.)

As a systemic problem, I’m not convinced that cars are the worst. Or outside what we accept in several areas.