Collier’s Weekly: Put Your Self-Driving Car Concerns in the Right Place

PHOTO BY KRISTY GRAVER

The city of Pittsburgh is about to confuse some robots.

Earlier this month, the autonomous car company Waymo announced it would deploy its self-driving taxis in a quartet of new cities: Philadelphia, St. Louis, Baltimore and Pittsburgh. The cars, already available in cities including Los Angeles and Phoenix, initially will be deployed with human drivers capable of overriding the car’s decisions.

I can only hope that these vehicles have some understanding of the city’s house rules.

Good luck interpreting a Pittsburgh left, robots! Hope you understand that the rules of engagement at certain chaotic intersections are more a matter of decorum and happenstance than those silly rules of the road! Speed limits? Variable according to how sensible we think they are! (Here’s looking at you, Bigelow.)

Although I think these vehicles are going to have a difficult time working things out around here, I am in no way opposed to their arrival. That goes against a certain strand of public opinion; there are those that firmly believe self-driving technology is too new and unproven to be tested on public roads. These machines are, inevitably, imperfect; there are numerous tales of the technology leading to catastrophe or dangerous situations.

Sure, the robots can make mistakes.

And if they make a whole hell of a lot of them, they’ll still be significantly better at operating vehicles than humans.

We flesh-and-blood drivers are generally pretty awful at safely moving cars around. More than 40,000 Americans died in motor-vehicle crashes in 2023, and 29% of those crashes involved speeding, an act that is typical for humans but strictly verboten for our autonomous counterparts.

That’s not the only bad behavior humans on which humans retain a monopoly. Among drivers in fatal 2023 crashes, 371 were using their cell phones at the time of the accident. Another 2,087 described themselves as “lost in thought.” More than 100 blamed their crash on a conversation with a passenger, and 37 said they crashed because they were distracted by food or drink.

Robots, as you may know, are particularly uninterested in food or drink.

Waymo, in fact, gathered heaps of data on the frequency of injury-causing accidents involving its vehicles. This data wasn’t purely internal; it was peer-reviewed and found to be entirely credible. Human drivers cause serious injury (or worse) more than 10 times as frequently as Waymo’s vehicles; even counting minor injuries, humans are five times as bad.

In other words: A bad robot is still a better driver than your average human.

We don’t fear driverless vehicles because they’re dangerous; compared to us, they’re not. We don’t trust them because we want someone to blame in the event of a crash. If calamity occurs, we want to make it someone’s fault; that’s difficult to do when there’s no face to ascribe to an accident.

There is, of course, a good reason to oppose the rise of driverless vehicles: The fact that they may put human drivers out of work. If Waymo and its competitors take over the taxi industry, it will mean cab drivers (and the hundreds of thousands of Americans who work for services such as Uber or Lyft) could be permanently out of work. In a just world, the cars would only mostly drive themselves, with human attendants remaining in the cars to navigate sticky situations, alter routes and generally tend to the needs of both driver and machine.

Is that likely to be the result of this shift? No — but the inevitable shrinking of labor demand due to the rise of advanced automation as a much more involved topic.

In the meantime, there’s no need for the arrogant opinion that we humans are safe drivers. We’re not. If you could snap your fingers and instantly make every vehicle autonomous, accidents and deaths on the road would drop precipitously. If we’re going to oppose this movement, let’s oppose it for reasons that are true — not the product of humans overestimating our own abilities.