This evening in New York, NPR broadcast a live press conference on the commuter train crash that killed four people in the Bronx. The crash has been on the news constantly, it's occupied page one of the New York Times website every day for the past three days.
Meanwhile, in the same period of time, I'm sure dozens of people have been killed in automobile traffic accidents in the New York metro area and there's been virtually no news about them.
This happens all the time and it points to one of they key problems that driverless cars will have to overcome: people's reaction to transit death is much more extreme when the driver wasn't one of us—i.e. an unpaid amateur, driving themselves.
It's not just the quantity of people who die in public transit crashes that make them news. It's the fact that none of the victims had control over their fate. It also doesn't matter how statistically better a bus or a train or a plane is. Traveling on a major airline is far safer than virtually any other mode of transit, including walking. But when people die on one everybody freaks out.
Now imagine what will happen when no one is the driver. A driverless car can be ten million times better than a human-driven one at avoiding accidents. It will get in an accident, though, and someone will die. And when that happens, the repercussions—especially those for the law and insurance—will be tremendous. Let's hope that doesn't slow their arrival down too much, though. We've been waiting for them for a long, long time—the picture above, an unfulfilled promise from 1957.