Confession: I hate driving. Some people enjoy it. I don’t. I never have. I don’t get the fascination. So, when it comes to self-driving cars, I may be a little (OK, a lot) biased. I want them yesterday, dammit. But I’ll try to be fair here. What’s good and bad about them?
Over the past several years, we have seen self-driving cars go from science fiction to being well within reach. You can’t buy your own just yet, but the technologies that enable them–adaptive cruise control, sensors that trigger brakes to avoid collisions, traction control technologies, and others–are available in cars today. Google plans to make their autonomous car model available as soon as 2020.
Self-driving cars are coming!
Unfortunately, when I see them discussed, there are a series of complaints and criticisms that tend to pop up. It’s not unfortunate that people are critical, but rather that the same vacuous complaints are made, over and over. I want to cover some of those first and just get them out of the way.
How can a computer drive better than a person? A computer can’t make better decisions than a real driver.
I hear that one a lot, and it’s not true. Everyone seems to think they’ve come up with some novel situation that a self-driving car couldn’t handle. While such a situation may exist, the bottom line is that no self-driving car has had an accident while driving itself, despite having over 2 million kilometers of self-driving distance logged. That’s more than most people drive in a lifetime, and rare is the person who has never had an accident.
The keys to this level of safety may be so obvious they are easy to miss. Computers can make decisions more quickly. They don’t get tired, or drunk, or angry. Perhaps most importantly, self-driving cars are programmed to always make the safest, most conservative decision. They don’t speed and attempt to overtake in a dangerous situation. They don’t drive too fast in bad weather. They don’t stop short in front of other cars out of anger. They don’t get distracted by their cell phones. I could go on.
It points to something a computer lacks: an ego. Perhaps one of the greatest failings of human drivers is our tendency to overestimate our own abilities. Having a realistic view of one’s competence is difficult. Automatons don’t have this sort of problem–they know what they know, and avoid making guesses about what they don’t.
What about car insurance? If a self-driving car malfunctions and causes an accident, who is responsible?
This one gets trotted out as a great big “gotcha!” even though there are multiple possible answers. It’s nowhere near being a dealbreaker. What’s true is that it’s unclear exactly how the presence of self-driving cars in significant numbers on our roads will affect insurance. What can be said already is that the medical costs of insurance claims are already on the decline, thanks to improved safety technologies in cars resulting in fewer deaths and serious injuries. Such trends are likely to continue as self-driving cars take to the roads.
So, who’s liable if a car is driving itself at the time of a crash? Perhaps one of the biggest advantages of a self-driving car is that it has lots of sensors and logging. There will be no ambiguity as to what the car was doing, or when, and it can be determined whether the car made the right decisions under the circumstances, or malfunctioned. Malfunctions would indicate liability on the part of the manufacturer, but that provides an incentive to make these cars as reliable as possible. After all, human-driven cars suffer manufacturing defects, as well, and those are handled via recalls, lawsuits, and government intervention. This may not be ideal, but the situation for self-driving cars is not likely to be novel. Car owners will still carry insurance, and either that insurance will be responsible for accidents, or the manufacturer will. Or, manufacturers will insure the cars themselves, and shoulder that burden. There are multiple possibilities, none of which are outlandish.
What might happen is that self-driving cars demonstrate themselves so much safer than human drivers that insurance rates for the latter become prohibitive, due to being so dangerous by comparison. That brings up the next question.
What about people who want to drive? What about people who enjoy driving?
First off, I think we are a very, very long way from an environment in which driving your own car on public roads is outlawed. What I believe will happen is that, at first, nothing will be any different. Most people will be driving their own cars, and self-driving cars will simply start to coexist with them. Even in cars still being driven by humans, safety features will develop to the point that it’s nearly impossible to have an accident, or at least difficult to have a serious one. Vehicles could be programmed to intervene in dangerous situations to avoid accidents and divert the car to safety. In effect, while a driver might have a semblance of control over the vehicle, the sensors and computers are still doing a lot of work to make it as safe as possible.
Eventually, we may come to a point where one must obtain a special license to drive a vehicle themselves, or to drive a vehicle that doesn’t have all of these safety features. Driving your own car on public roads may become difficult, from a regulatory perspective. Private tracks and roads may become more popular as people who enjoy driving as a hobby find themselves in need of such an outlet. But there’s a good chance that, after a few decades, the option to drive your own car will be out of reach for most people. To be blunt about it, I don’t think anyone’s hobby is worth tens of thousands of deaths a year.
What about jobs? A lot of people drive for a living. Won’t they be automated out of work?
I can’t say that no jobs will be eliminated by self-driving cars. It’s likely that many will be. The US has about 3.5 million professional truck drivers. It’s hard to say how many might see their jobs vanish as trucks become capable of driving themselves. There will almost certainly need to be human beings in the cab for the foreseeable future, if only to monitor the vehicle and handle emergencies, breakdowns, and maintenance. They may no longer spend their time actually driving the truck, just making sure it reaches its destination without incident. The question then becomes whether a driver’s compensation goes up or down. I would argue that it could go up, if a driver’s responsibilities shift more toward technical skills rather vehicle operation. Only time will tell. But I think it’s also fruitless to object to technologies that make jobs redundant–the discussion should be about what to do with people whose jobs have been eliminated, not how to keep them from being eliminated in the first place.
On that note, one job I am less optimistic can be preserved meaningfully is that of a taxi driver. Moving people over relatively short distances within a well-mapped urban environment strikes me as something that could be automated a lot more easily than some other positions. Taxis would still exist, but taxi drivers may soon be obsolete. Taxis could become another arm of public transportation in many areas, serving short legs of urban journeys that aren’t covered by buses, trams, or subways.
Overall, I think self-driving cars will prosper primarily due to safety and convenience. Cars that can drive themselves open up new possibilities, such as pooled ownership in which several families could share a car and pass it between each other with ease, as needed. I could see there being a monthly service that has such a fleet on hand to deliver a car to you, as needed, any time you like–imagine Zipcar meets Enterprise, but fully automated. The safety aspects are very impressive, as well, and I suspect that once the general public has a better sense for just how much safer and more convenient these cars are, people will clamor for them and snap them up as they come down in price. The technology for vehicle automation is already reasonably priced for what it is, and sure to come down as it becomes more prolific and refined. I think cars are in for an interesting future, and I can’t wait for it to get here.
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.