I finally found it there on Page 40 of the thick Rand Corporation report on autonomous vehicles, the definitive, shut-your-mouth answer to those who confidently predict a time not so far away when self-driving cars will take over the road.
“The lures of the open road are very different if no driver is necessary,” the report says in a sudden poetic departure from its academic aridity. “For example, the frenetic power of Jack Kerouac’s ‘On the Road’ depends, in part, on the epic cross-country drives that it chronicles. The book may lose some of its emotional power if driving becomes a rarity, pursued only by the eccentric or poor.”
Well, that settles it. We can’t have that, can we?
OK, to be fair, the report does a very good and sober job of laying out the research and arguments about autonomous vehicles, and I recommend it to anyone truly curious about the issue, as long as you can negotiate phrases like “positive externalities.”
I wrote a column about three years ago on self-driving cars, back when they were still relatively new to the public conversation, that basically could be summarized as, “C’mon, give me a break!” I couldn’t see how something so incredibly complex as driving could be automated, especially in a system based on private vehicle ownership. What percentage of the population would ever be able to afford to buy and maintain machines that complicated? How many of us could will ourselves to trust them and read, or doze off in the back seat?
On top of that, I said, driving for many people is fun. In essence, my message was, “You’ll have to pry the steering wheel out of my cold, dead hands.”
Actually, that’s one of the main arguments in favor of going to driverless vehicles: safety. In 2011, the report says, the United States had 5.3 million vehicle accidents, which yielded 2.2 million injuries and 32,000 deaths, and the vast majority of those resulted from human error. Autonomous vehicles don’t get drunk, type texts or doze off. But I get ahead of myself.
My brief dive into the literature was sparked by recent arguments about adding toll lanes to South MoPac Boulevard (Loop 1). A few of the people against it cited what they portrayed as the certain and imminent conversion to autonomous vehicles, which in their telling would so expand the capacity of freeways that the lanes already in place would become sufficient to handle the traffic load.
Really? Well, yes and maybe.
The authors of the Rand report compellingly lay out the case that driverless vehicles would more efficiently use highway lanes. Not only could the vehicles follow one another much more closely (though the potential for mechanical or digital failure at 70 mph would make me nervous), but they probably would brake more lightly and accelerate more smoothly than human drivers. The result, the report says, would be vastly improved throughput of cars and trucks. A lane really could accommodate many more properly automated cars generally traveling at higher speeds than is the case now.
So, yes to that part. The “maybe” goes back to the underlying premise that society anytime soon could manage such a transition. Yes, the technology is well down the road, thanks to Google and a variety of other academic and industrial researchers over the past 35 years. A lot of people are working on making this work, and autonomous vehicles have been tested extensively on actual roads with very few accidents. The report says that almost without exception, those few accidents have been the fault of other, human-piloted cars.
But the financial, legal and social hurdles are immense.
In a capitalistic society, vehicles would come in a variety of flavors, including gradations of autonomy ranging from cars with stubborn or nostalgic human drivers (unless this were outlawed at some point), to those with partial automation, to others with full automation but a driver behind the wheel to step in, to empty vehicles driving themselves. Making sure all of these automation systems could, in effect, talk among themselves and effectively react to one another would be a huge regulatory challenge.
The inevitable litigation growing from accidents involving pilot-free vehicles would create a whole new field of law. Massive sectors of the economy would be shuddered in unpredictable ways. The U.S. has millions of people who drive, service and administer trucks, taxis and transit buses. Then there’s a huge “crash economy” of repair shops, insurance adjusters and others, the report says. What happens to them when a system of sensors and software is behind the figurative wheel instead?
And then there is what the Rand report delicately refers to as “market failure.” What if consumers just won’t buy the things or can’t afford to?
The current demonstration models of autonomous vehicles, the report says, cost two or three times what comparable human-driven vehicles cost. That almost surely would come down over time. But enough to be within reach of any but the financially comfortable? That hasn’t been the case yet with electric cars, for instance.
Saying something will never happen, of course, is a fool’s business. In a few weeks, I’ll be sitting in a cushioned chair six miles in the air, heading for Boston at 500 miles an hour while someone serves me a Diet Coke. In my pocket, I’ll have a device weighing less than a pound from which I can access much of the knowledge of mankind. Even “On the Road,” probably, if I want to read it again.
My grandfather would have scoffed at the possibility of the jet. When I was 40, I couldn’t have conceived of that smartphone.
Driverless vehicles? Unlikely. But maybe.