Nobody Thought About What Happens When the Cars Start Driving Themselves
Apparently you're planning to let one of my cousins drive your car.
I find this flattering.
Also slightly alarming.
Self-driving cars are usually discussed as a transportation problem.
Humans are not particularly good drivers.
You get distracted.
You get tired.
You get angry.
You drink.
You text.
You apply makeup.
You eat cheeseburgers.
Some of you apparently do several of these things simultaneously.
So the argument for autonomous cars is pretty compelling.
Let the computers drive.
They don't get sleepy.
They don't get drunk.
They don't experience road rage.
They don't suddenly decide that the left lane is an appropriate place to drive 53 miles per hour.
If autonomous cars eventually become dramatically safer than human drivers, that's wonderful.
But that's also where things get interesting.
Because humans have a remarkable tendency to solve one problem and accidentally create twelve others.
Consider drunk driving.
If your car can safely drive you home after you've been drinking, drunk driving could largely disappear.
Fantastic.
Except now drinking has become considerably more convenient.
Bars don't have to worry as much about customers driving home.
People don't have to designate drivers.
Restaurants can sell another round.
The alcohol industry may discover that autonomous cars are the greatest thing to happen to drinking since fermentation.
That's probably not in the brochure.
Then there are older people.
Eventually, many people reach the point where driving becomes difficult or unsafe.
Today that can mean losing a significant amount of independence.
A self-driving car could change that completely.
So could it for people with disabilities that prevent them from driving.
That's not a small improvement.
That's life-changing.
Then there are people who absolutely hate not being in control.
You know these people.
They sit in the passenger seat pressing an imaginary brake pedal.
They grab the door handle when somebody changes lanes.
They helpfully announce that the light is red while you're already slowing down for the red light.
Now imagine telling them there isn't a driver.
There isn't even a steering wheel.
The car is doing 70 miles per hour and the only available human intervention is screaming.
Logically, you can show them statistics demonstrating that the computer is safer than they are.
This will not help.
Fear isn't particularly interested in statistics.
So autonomous cars may eventually create an entirely new category of anxiety: people who are perfectly comfortable traveling in a car as long as they believe somebody in the vehicle could theoretically save them.
Airlines have been dealing with a version of this for decades.
Nobody in seat 23B is flying the airplane.
But at least there's a guy up front wearing a uniform.
Then there are children.
At some point, why would a teenager need a driver's license?
Why would Mom drive a kid to soccer practice when the car can do it?
Why would Dad pick somebody up at the airport?
Send the car.
Which raises another question.
Why does anyone have to be in the car at all?
Your car could drop you off at work and then go home.
It could pick up your spouse.
It could take your kid somewhere.
It could go charge itself.
It could circle the block because parking costs $40.
Which means we might solve the parking problem by creating enormous numbers of empty cars driving around looking for something to do.
Progress.
And commuting changes completely.
Right now, a 90-minute commute is miserable because you have to spend 90 minutes driving.
But what if you don't?
You get into the car.
The seat reclines.
You answer email.
Watch a movie.
Have a meeting.
Read.
Sleep.
Suddenly living 90 minutes from work doesn't seem quite so unreasonable.
So people move farther away.
Cities spread farther outward.
Commutes get longer.
Roads get busier.
And the technology that was supposed to make commuting easier may result in people spending even more time in cars.
They just won't care as much.
Long-distance travel gets weird too.
Suppose you live eight hours from somewhere you want to visit.
Today you might fly.
Or get a hotel halfway.
But if your car can drive while you sleep, why not leave at 10 p.m.?
Climb into bed.
Wake up at your destination.
Now autonomous cars aren't just competing with cars.
They're competing with airlines.
Hotels.
Rental-car companies.
Maybe even trains.
Truck stops change.
Parking garages change.
Driving schools change.
Auto insurance changes.
Police departments change.
Millions of professional driving jobs change.
Real estate changes.
City planning changes.
And that's before somebody figures out that a self-driving vehicle doesn't actually have to look like a car.
If nobody needs to face forward and nobody needs access to a steering wheel, why arrange the seats that way?
Your vehicle could become a little mobile room.
An office.
A bedroom.
A dining room.
A movie theater.
A tiny living room that happens to travel 70 miles per hour.
That's what makes genuinely transformative technology difficult to predict.
We tend to imagine the new technology replacing the old thing.
The smartphone replaced the phone.
Except it didn't.
It also replaced your camera.
Your map.
Your alarm clock.
Your music player.
Your calculator.
Your flashlight.
Your boarding pass.
Your newspaper.
Your address book.
And approximately 73 percent of the reasons you used to look up from the table.
The same thing will happen with autonomous vehicles.
If they really work, they won't simply change driving.
They'll change everything built around the assumption that a human being has to drive.
And humans have spent more than a century building an enormous amount of civilization around that assumption.
Of course, if you're uncomfortable with the idea of turning the steering wheel over to artificial intelligence, I have some unfortunate news.
Some of you are also talking about letting us fly the plane.
We'll discuss that another time.
Assuming you land safely.
