The Friction and the Fortress

Why good innovations mint new elites — and how to tell the dangerous cases from the merely disruptive Every year something like forty thousand Americans die on the roads. The strongest large-scale data we have — from driverless fleets logging hundreds of millions of miles — suggests that within the limited areas where they operate, robotaxis injure people far less often than human drivers do. The data isn’t fully independent, and the operating areas are cherry-picked for ease. But the direction is not really in dispute. So who lobbies hardest against letting these cars spread? Among the most persistent opponents, for nearly a decade now, have been trial lawyers. The arithmetic is not subtle. Ordinary auto litigation is a vast industry, and an industry built on collisions does not welcome a technology that removes most of the collisions. This looks like a clean morality tale: lifesaving progress on one side, a rent-seeking guild on the other. It is a clean tale. It is also the weakest…