James Pethokoukis:

Three thoughts here. First, there is huge potential upside. With high penetration rates, research suggests, driverless technology could save 22,000 lives a year and $450 billion annually.
 
 Second, one could also see a “bootleggers and Baptists” scenario develop with public-minded interest groups (maybe robo-apocalypse worriers or “for the children” types) allying with cronyists or rent seekers (this massively disruptive a technology is going to make some incumbent somewhere unhappy) to bring down the heavy hand of government regulation.
 
 Third, there is a big difference between driverless cars where can you take naps or immerse yourself in a book and near-driverless cars where the driver must remain alert and ready to act. But humans aren’t very good at such passive observation. This from a New Yorker piece on challenges posed to airlines by automation and growing pilot inattention: