I look forward to the singularity—the moment when artificial superintelligence triggers an explosion of technological growth that leaves us humans behind. Or, more specifically, I look forward to the day sentient computers are driving your cars. Because, by and large, you’re terrible at doing it yourselves.
Don’t get me wrong, humans excel at a broad range of important tasks; this includes the creation of life, and the morality, art, philosophy, and institutions that make that life worth living. You are the inventors of wondrous technological advances that have allowed us to experience longer and freer lives. Yet surely we can agree that humans are also hampered by limitations. Have I mentioned driving?
According to empirical data compiled over 20-plus years negotiating America’s traffic, I estimate that there’s an 88 percent probability that you’re a road hazard. You’re probably indecisive and impatient. You weave, text, and make unconscionable U-turns and perilous lane changes; you needlessly tailgate and flicker your brights; you drive lethargically in the fast lanes and at hazardous speeds in the slow ones; you attempt head-first parallel parking and experience outbreaks of uncontrollable road rage when things aren’t going your way; and, worst of all, you rubberneck.
Driving up and down the East Coast every holiday season has convinced me of one thing: The day of automated self-driving cars couldn’t possibly come quickly enough.
This position is not born wholly of selfish motives. I beseech you to think about our beloved septuagenarians and octogenarians as they risk their lives (and those of all of us) ambling down our nation’s byways in two-ton death machines at 20 miles per hour. Think of the parents who are forced to operate heavy machinery while unruly children distract them with endless questions and whining. Think of our handicapped; the deaf, the blind, and teens. We will soon have the technology to help them.
Around 30,000 people perish yearly in vehicular accidents. When we consider the drop in deaths and the increase in vehicle miles driven per person over the past five decades (not to mention how poorly we negotiate our roads), it’s clear we’ve made tremendous strides in safety.
Still, something like 95 percent of all motor-vehicle crashes are caused at least in part by human error. Automated cars could save tens of thousands of lives each year. A study by Department of Transportation claims that robots could also potentially save the nation around $160 billion a year on gas and time gained by eliminating traffic. And what right-thinking person wouldn’t answer to a robot overlord if it meant eliminating rush-hour traffic?
Many, apparently. One recent poll found that only 23 percent of Baby Boomers would ever have faith in self-driving technology. Younger people do tend to be more accepting of automation. But no one is sold on the prospects completely. Around 41 percent of Gen Xers would trust robot cars and around 56 percent of Millennials.
In a round-up of national attitudes on the issue, one automotive magazine proclaimed that those who have “endured computer crashes as part of their everyday existence are wary of trusting software to keep them safe.” Have these people ever met their neighbor? For that matter, have they ever flown in an airplane or taken a train? Both methods of travel use computers to circumvent human error.
Now, nearly every time I praise the glorious driverless, traffic-less future, I’m faced with similar pushback: Would you entrust your family’s welfare to a machine? Of course not. I’m a fantastic driver. This isn’t about me. What’s important is that I would trust computers more than I trust you on the same road as my loved ones. Computers, I’m afraid, are far more reliable than people. They tend not to drink and drive or fiddle around with the radio or put on makeup on their way to work.
Of course, I’m no safety fascist, either. A bubble-wrapped life is a boring one. Nor am I a techno-utopian. I certainly detest social engineering. I don’t believe anyone should be forced or nudged to participate in this driverless world. If the accessibility, reliability, and safety of driverless cars shakes out as expected, the market will make switching feasible and attractive.
Google, a leading innovator of autonomous-vehicle technology—though nearly every major car manufacturer is pushing forward on this front—recently marked more than 2 million miles driven on public roads without an accident of any kind.
I’m skeptical that robots will ever possess self-awareness in the way a human experiences it. Machines tend to be programmable, subservient, and useful. Isaac Asimov’s famous “Three Laws of Robotics” seem like a perfectly serviceable set of ideals for our cars to follow. (1. Robots may not injure human beings or, through inaction, allow human beings to come to any harm. 2. A robot must obey orders given it by human beings except where such orders would conflict with the First Law. 3. A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.)
The Office of Naval Research has also awarded $7.5 million in grants to researchers at top schools such as Brown, Georgetown, Tufts, Yale, and Rensselaer Polytechnic Institute to study the possibilities of teaching robots the difference between right and wrong and the weight of moral consequences. I find this heartening rather than frightening, because you never know when a car will have to make an ethical choice—such as, say, decide whether to run over a person or a deer.
What I do know, though, is that if my Honda becomes a conscious being, I would still trust it more than most I-95 drivers on Black Friday.
– Mr. Harsanyi is a senior editor of the Federalist.