Early this year, the wife and I became parents. Having become parents, we found ourselves (or I have found myself; I shouldn’t necessarily implicate the wife in my fretting) stressing out about all the various future situations in which our daughter might be in danger. Car accidents are up there near the top of the list of worries. I was the worst sort of over-aggressive driver as a teen, with the body shop bills to prove it, but I managed to escape from that period of my life without doing myself, or anyone else, bodily harm. I consider myself lucky. And while I take some solace in the fact that young females may not necessarily be as idiotic behind the wheel (insurers charge more for males, so I defend my gender stereotyping by referencing the actuarial tables), I understand that teenage females occasionally ride in cars with teenage boys. Tens of thousands of Americans are killed each year in automobile accidents, a disproportionate share of them teens. No one thinks their kid will be among them, but thousands of kids are, every year.
I was thinking on this this morning, when I realized that there’s a decent chance my little girl may never need to learn to drive. And not just because she’s opted for the walkable, transit-friendly lifestyle her dad’s always going on about. By the time she reaches legal driving age, I thought, autonomous cars may be common on American streets.
I tweeted this thought and received some pushback from others, including Tim Lee, who’s given the possibility of widespread adoption of autonomous cars some serious thought. We debated the topic a bit on Twitter and Tim ultimately offered me a bet:
I bet you $500 that on your daughterâ€™s 16th birthday, it wonâ€™t be possible and legal for someone with no driverâ€™s license to hop into a self-driving car in DC, give it an address in Philly, take a nap, and wake up at her destination 3-4 hours later (depending on traffic, obviously).
The car must be generally commercially availableâ€“not a research prototype or limited regulatory trial. It can be either purchased or a rented â€œtaxi.â€ And obviously there canâ€™t be anyone helping to guide the vehicle either in the car or over the air.
Which I have accepted. Tom Lee promises to referee the bet, and you, readers, are now witnesses. Tim has set out some of the thoughts underlying his position at his blog (linked above). I’ll briefly summarize my own thinking now.
The technology for reasonably reliable autonomous cars seems well within reach. Cars already do a lot more driving for us than most of us realize, and others (including Google and the Department of Defense) are testing full-on autonomous vehicles. The big constraint on widespread adoption of robocars within 16 years doesn’t seem to be technological. Rather, it’s all about human comfort levels and the legal and regulatory environment. Tim makes some strong points on this count, and I wouldn’t be surprised if he turns out to be right. But I have a sense that things might just fall into place for robocars.
Why? Several things. First, the opportunity cost of time is steadily increasing. Thanks to huge improvements in mobile communications, people can do all sorts of things while outside of their homes and offices, and that makes time spent sitting at the wheel of a car focusing one’s attention on the road increasingly costly. People attempt to get around this by using devices while driving, but that’s dangerous and authorities are going to ever greater lengths to prohibit such activity. Mobile devices improve the cost profile for transit, but it’s difficult to imagine transit reaching a much larger share of American households within two decades. If anything, underinvestment in infrastructure over the next few years seems likely to combine with population growth to worsen congestion problems and lengthen commute times, thus making driver-operated vehicles more of a pain.
So working people will be anxious for the freedom to work granted by robocars. Young people, who’ll be as addicted to technology as ever, may be unwilling to set foot in a car that doesn’t allow them to tune out and immerse themselves in their electronics. The market will favor robocars.
Meanwhile, I think people might be more accepting of robocars for safety reasons than you might think. I think you see a bit of this in the taste for a growing array of automatic safety features on automobiles, from devices that help keep you in your lane or which brake automatically when the car in front of you stops short, to simpler things like airbags and automatic traction control. GPS isn’t entirely unrelated. People are remarkably willing to surrender themselves to automotive technology already. Robocars don’t strike me as that big a chasm to leap.
Particularly among groups at risk for bad driving. Parents unwilling to turn their own lives over to a robot might be more than happy to let a computer take the wheel rather than their teenage son. The elderly may opt for robocars in droves, if it means continued mobility. The same may be true for the disabled. Or those who like to drink. When you start adding up all the various constituencies for robodriving, you end up including quite a large share of the population.
And insurance goes both ways on this. Anyone basing their driver choice decisions on raw statistics won’t need to see much of an improvement over human drivers to begin flocking to autonomous vehicles. Insurers may offer big discounts for families who use such cars. Fleets may be anxious to take the job of driving out of the hands of fallible human workers.
Ultimately, it’s not hard to see interests lining up to support a relatively liberal approach to autonomous vehicles. If auto-driving technology continues to ease its way in, 16 years to robocars seems like a safe bet. if its more of a single-hurdle legal process, rather than an evolutionary change, that will make things harder. But it’s also the kind of thing where once the technology proves itself, it may be hard to resist, given that demand is likely to be high, and there aren’t huge interest groups standing to lose from the shift. Where there are such groups, exceptions may be made. I wouldn’t be surprised if trucking was still human-dominated in 16 years. But personal transportation? I’ll bet that we’ve fired ourselves as drivers by 2026.