The Press Box - When the Robots Reach the Road (Ep. 308)
Episode Date: May 19, 2017The Ringer's Ben Lindbergh, who still doesn't have a driver's license, investigates how far away we are from a driverless future, the utopian (and dystopian) visions of what a driverless world could l...ook like, the hurdles that have to be cleared, and whether driverless cars can arrive in time to save him (and other licenseless citizens) the trouble of ever learning to parallel park. Learn more about your ad choices. Visit podcastchoices.com/adchoices
Transcript
Discussion (0)
I recently turned 30. Unlike anyone who hits that milestone, I asked myself some uncomfortable questions.
Does this mean I'm past my prime? Or my musical tastes now frozen forever? Is it time to get a driver's license?
Actually, that last one might sound sort of strange. But I still don't drive. I don't even have a license.
On a recent Saturday, I walked around my Manhattan neighborhood, looking for fellow driving holdouts.
For the most part, I found people who did have licenses, but who weren't totally sure why.
I have a driver's license, but I haven't driven since I was 18.
Okay.
I actually also didn't renew it, so I might not have a driver's license.
So why did you get one?
I got one because my parents made me.
I was like, why would I get a driver's license?
And they were like, well, we're not going to drive you around all the time.
And I was like, just never used it because I'm in the city.
Do you have a license?
I do.
I don't regret getting one.
Everyone should be able to get their driver's license at some point in their life.
I just think that for me personally in New York City and it's really like kind of a metropolitan city,
it's just like not worth it for me.
So we're just wondering if you guys have driver's licenses?
Yes, I have a driver's license.
I have a driver's license.
Yes, I do.
You all do.
So how often do you use them?
Like almost never.
Actually use my driver's license?
Like never.
Do you have a license?
I do.
And how often do you use it?
Not very much.
I haven't moved my car in three weeks, four weeks.
There are two reasons why I haven't given in.
First, I grew up in Manhattan.
and so I never needed to drive.
Most things were within walking distance, and whenever something wasn't,
there was always a bus or a train or a taxi that could take me where I wanted to go.
In Manhattan, owning a car either costs a fortune or leads to a lot of time spent searching for parking.
Which brings me to the second reason I don't drive.
The Masher.
That was the name of our family car when I was a kid,
a 1978 Pontiac Lamans that was way older than I was.
We bought it from a friend for $200.
It didn't lock and had no working clock.
The radio was only AM,
and the ceilings upholstery was so exposed,
that tiny bits of it would sprinkle on us while we sat, like ferry dust or car dandruff.
We parked it on the street, where every few days it would collect a new dent or ding.
One day, we found it with the whole driver's side bashed in.
The masher had been sacrificed to the streets of New York.
Most days, it doesn't matter that I don't drive.
But when I'm outside of the city, it makes me a burden.
Two years ago, my girlfriend, Jesse, and I spent the summer in Sonoma while I was working
on a book.
Because she had a license, Jesse became our de facto chauffeur.
So how much time do you remember spending driving me around Sonoma?
I drove you around every single day, and they were long drives.
Okay.
And how much did you enjoy driving me around Sonoma?
In general, I would say I did not super enjoy it a whole bunch.
And how do you feel about the fact that I can't drive?
I understand it.
I see why that's the case.
I'm empathetic to the non-desire to get a driver's license.
Well, and also, even if you had a driver's license, I don't know why you would use it, because you never leave the apartment.
So, no, I understand it's fine.
For some reason, she agreed to marry me anyway, but I still feel like a freeloader when I can't take a turn on long trips.
And sometimes I think I should get a license just so my mom will stop telling me to.
Yeah, well, what if I'm driving upstate and I get tired and you had to drive and you don't have a license?
So, have you scheduled your road test yet?
I do have a learner's permit.
Two years ago, I finally found out what all those mysterious signs and lines meant and took the written test.
But I could count the number of times I've been behind the wheel on one hand.
Manhattan made me think driving was unnecessary, and the masher made me think cars weren't cool.
But driverless cars are cool, and they're coming.
If you haven't seen or ridden in a driverless car, you've certainly read about them.
We're heading for a future where no one will drive.
The question is how far away that future is.
If driverless cars come quickly enough, my procrastination never needs to end.
I can keep walking, taking public transportation, or bumming rides from friends and family until the robots bail me out.
So that's what we'll find out on this episode.
How close are we really?
to truly autonomous vehicles.
What will the driverless world look like?
And based on what we know now,
should I finally learn to parallel park
or accept that I'm never going to try?
I'm Ben Lindberg, a writer for the ringer.com,
and you're listening to a very special tech episode
of Channel 33.
I'm not alone in forsaking the car.
In my age group, it's becoming more common.
We are seeing trends in the industry here
where car ownership is going down.
The number of people getting licenses is also going down.
That's Taggart Matieson,
the director of product for the ride-sharing company Lyft,
which just partnered with Waymo, the driverless car developer owned by Google's parent company.
Matiesin says that while I've been pumping the brakes on driving, driverless technology has accelerated.
We're really kind of at the cusp of a number of technologies coming together
where we can actually move from our current network to a network that supports self-driving vehicles.
Until recently, it was almost impossible to live in a lot of places without owning a car.
But today, ride-sharing apps have made it possible to live even in decentralized cities like Los Angeles,
without driving. I might be wrestling with the drive or don't drive dilemma a little earlier than most
Americans, but before long, everyone will be. The first thing you have to know about autonomous
vehicles is that most of them aren't truly autonomous. They still require some level of human input
and consequently, a license, which makes them off limits to me. As my colleague Victor Lukerson,
who covers technology for the ringer and rubs it in my face that he got a license at 16, explains.
The Society of Automotive Engineers has created this ranking of autonomy for vehicles,
going from level zero to level five.
Level zero is basically every car from the Model T until maybe like a decade ago.
The next lowest level, level one, or as SAE calls it, driver assistance, probably describes your current car.
At this level, the car can help out with one task, including cruise control, but it can't do anything more complicated than that.
Level two is known as partial automation, and in that case, a car could, for instance, automatically know when to change lanes
and know when to accelerate or brake to stay within traffic.
Tesla right now, with their autopilot function, does this.
Level three is conditional automation.
At level three, a car could make decisions.
For example, instead of just braking to avoid another vehicle, it could decide to switch lanes and pass it.
No one has brought a level three car to market.
The car is highly capable.
Theoretically, the humans should be able to relax a little bit more.
But at the end of the day, in the event of an emergency, you still need to be able to grab the wheel and take control of the vehicle.
Okay.
Doesn't work for me.
Next.
All right.
Level four, which I think this one might be your ticket, is known as high automation,
which means that there's no human input involved at all.
Perfect.
And the car doesn't even need to have a wheel or a brake pedal.
Okay.
So I feel like if you can find yourself a level four car without a steering wheel or a brake pedal,
then you might be able to maneuver that one.
The only problem with level four vehicles is that they're limited to environments
for which we have very detailed and accurate digitized maps.
That might work in Manhattan, but if I want to go upstate for the weekend
or have to travel to a rural area out of state, I'm going to need totally autonomous tech to go all the way.
Level 5, full automation would be a situation where you have a driverless car, the does require human input, and can maneuver any road at any time, which would either require an incredibly detailed map of all roads in the United States or sort of a level of AI intelligence that we don't really have yet where the car could understand and manipulate an environment the way the human mind can.
If you want to be able to go wherever you want at all times, you're going to need a transformer essentially.
AutoBOTS. Transform and roll out.
Okay, so who's leading in the transformer building business?
The company's on the cutting edge are a mix of Silicon Valley in Detroit.
Google has done the most research.
This year, their self-driving car project, which was created in 2009,
reached a milestone of 3 million miles covered, mostly in California.
Google subscribes to more of a revolutionary approach than an evolutionary one.
They think autonomous vehicles will be safer if we strip away the steering wheels,
removing all temptation for us flawed, fallible humans to,
intervene, Google engineers are designing a user interface that will put passengers at ease with their
lack of control, but the company has yet to reveal its plans to monetize its technology.
Their technology is thought to be the best, but they've also been one of the more cautious
companies. They have yet to launch a commercial product with their driverless tech, and their
tests have been relatively constrained. If you're talking about who's been the most aggressive
and been the first to market to customers, that would be Uber. They're currently testing
actual driverless cars in Pittsburgh and Phoenix with real customers. So if you go to one of those two
cities, you can get in a driverless car right now.
Although the venture capital arm of Google's parent company, Alphabet, invested more than
$250 million in Uber in 2013, the two companies' autonomous arms race has strained their
relationship, leading to an alphabet executive resigning from Uber's board, rampant poaching
of talent, Waymo's partnership with Lyft, and even the pursuit of injunctions against Uber's
driverless development based on allegations that Uber use secrets stolen from Waymo to advance
its own efforts. It's possible that while the two tech giants are squabbling, they'll be passed by a
company who's founding predated Googles by 95 years and Uber's by more than a century.
In terms of actually committing to a mass product, Ford is probably in the lead of the pack.
They've committed to by 2021 having a mass production vehicle with no steering wheel and no brake pedal.
Ford is the one I would say is been the most aggressive about committing to a real timeline,
and that's only four years from now.
And when it comes to any mode of futuristic transportation, Elon Musk and Tesla must always be mentioned.
With their autopilot software, they're able to collect a lot of data right now,
all the roads that Tesla's are driving on.
They actually updated their vehicles last year
so that any new Tesla you buy today
will eventually have self-driving software enabled.
At some point in the next couple of years,
Tesla will be able to flip a switch and make their car self-driving.
We just don't know exactly when that's going to be yet.
We don't yet know which of those companies,
or their competitors, including Lyft, GM, and Apple,
will make the most money from driverless vehicles.
But Mark Hallenbeck, the director of the Washington State Transportation Center
at the University of Washington,
confirms that the technology is most of the way there.
It's just that the last step of development is going to be a doozy.
In engineering terms, we like to say that the first 95% of the development of any new technology takes about 5% of the cost.
And the last 5% takes 95% of the cost.
And that's kind of where we are in the automated vehicle world.
We've done the technical stuff.
We haven't done the weird exceptions stuff.
but when you're expecting the car to deal with those weird exceptions at 70 miles an hour,
and they don't, you die.
One of the innovators pushing driverless technology forward from today's level one and two cars to tomorrow's level five is an Israeli company called Mobile Eye.
Once partnered with Tesla, MobileI, which was bought by Intel in March for more than $15 billion,
is now working with BMW and Volkswagen to develop autonomous vehicles for consumers by 2021.
Yeah, I'm Dan Galves.
My title is Chief Communications Officer. MobileI developed software takes raw camera data and turns it into
usable information that tells a car basically what its surroundings are.
Today, that information is used in collision avoidance systems that can activate a car's brakes in the case of an impending impact
or warn a driver who's drifting out of their lane. In the coming years, it could completely control the car,
allowing its human passengers to take a back seat, both figuratively and literally.
In order to do that, you need to be able to see a wide,
field of vision. So we have a product coming out that has kind of three lenses in the front where
one lens is very wide angle, one is narrow and long distance and a mid-range. You'll also need
redundancies. Once you tell the driver, you don't need to pay attention anymore, you can't risk
the camera being disabled or not working properly. So you'll start to see radar sensors or
LIDAR sensors that can also identify objects around the vehicle and basically double check what the
camera's seeing.
To get from there to the level 3, 4, and 5 future where I'll finally fit in, though, are super-perceptive
driverless vehicles will need to know more about where in the world they are.
GPS map really only can tell the car where it is within 5 to 10 meters.
The car needs to know where it is relative to the boundaries of the road with a lot more
accuracy than that.
Galv says that the technology for the sensing part of the problem is pretty much ready.
It's the mapping part that remains a major logistical problem.
Someone has to build a high-definition map that doesn't yet exist.
The conventional approach to building these maps, to the extent that any of this is conventional,
would be to attach specialized equipment to dedicated vehicles and hire people to drive them around,
as Google has done with its street-view cars.
With enough time and money, you could build a map that way, but you couldn't keep it updated,
and you can't have autonomous vehicles operating based on out-of-date maps.
Fortunately for the future of driverless vehicles,
and unfortunately for people with privacy fears, cars are already crowdsourcing their surroundings.
Smile, you're on collision avoidance camera.
We have cameras going on a whole lot of cars today to do these collision avoidance systems.
So these are not autonomous cars.
They're just regular vehicles driving around, but we have a technology that can essentially
pick out stationary landmarks and take measurements of where the car is passing those landmarks
and build a very simple map that way.
but it's a map that then autonomous cars can use to know where they are in the world to a 10 centimeter accuracy level.
And since these are millions of just regular vehicles driving around,
when they pass a part of the world where what the camera is seeing is different from what the map says it should see,
then they can make an update to the map.
Ironically, once we have superhuman driverless cars that are more observant than we are and better navigators than we are,
we have to teach them to behave more like we do.
Essentially, we have to teach them to take risks.
The initial development of autonomous cars would lead to very conservative vehicles,
where you would always wait for the perfect opportunity to merge.
With new artificial intelligence techniques, you can simulate and train these systems
through reinforcement learning, through rewarding good behavior and punishing bad behavior
in a virtual reality environment.
You can train them to have some of these human driving skills and actually drive more aggressively.
One team led by Alan Cornhouse,
a Princeton professor of operations research and financial engineering,
is trying to accelerate the process by training AI on virtual environments found in video games.
When we, well, when you drive a car, you learn responses to certain stimuli.
For example, you see a stop sign ahead and you know to hit the brakes at the proper time
and come to a complete stop, especially if someone else is watching.
An algorithm or AI can undertake the same learning process,
but as Kornhuser says it needs images to train on.
You have really two choices.
you can put a camera on a dash of a car, drive around,
and then go back and look at each one of the images
and label them yourself or send them to a crew of labelers
somewhere around the world and have them label it.
The problem with that is, of course, it's very labor-intensive,
and it's not necessarily correct
because someone might look at a stop sign
and not recognize it as a stop sign,
or think that the stop sign is 43 meters ahead
when it's actually 38 meters ahead.
The challenge Cornhouser says is to get images that cover every set of situations you can expect
while driving and have them all be pristine.
A virtual environment is perfect because its creator has all the answers.
Cornhouser's team used a racing game to train an AI to drive safely on virtual courses
it had already been trained on, and subsequently on virtual courses it had never seen.
Obviously, though, a racing track is different from an everyday driving situation,
which is why the fine print disclaimers on those devil-may-care car ads say,
professional driver on closed course and tell you not to attempt what you see on the screen.
For a more realistic test, Kornhouser has moved on to schooling algorithms with a driving
game that tries to simulate a real city, Grand Theft Auto 5.
If you saw me driving in Grand Theft Auto, you definitely wouldn't want me driving on a real-life road.
But GTA might help ensure that I won't have to.
I mean, there are pedestrians, there are real street scenes, there are stop signs,
there are traffic lights.
And the other piece of it is that you can control the weather.
You can have it rain, you can have fog, you can have snow, and it's fairly realistic.
We went in there and have found that, in fact, it is a very rich environment.
And this whole concept of using virtual reality as a training, educating environment
for deep learning neural networks to drive cars safely might actually have legs.
So that's where the technology stands.
But there are other hurdles that will have to be cleared.
For me to do without a driver's license, it's not enough for autonomous vehicles to be feasible.
I also need them to be allowed on the roads.
For the latest on the regulatory timeline, I talked to Bryant Walker Smith, a member of the legal faculty at the University of South Carolina,
who was recently appointed to the Federal Advisory Committee on Automation and Transportation.
Everyone's talking about the need to legalize automated driving.
But is that the wrong perspective?
Should we actually be asking whether automated driving is legal now?
In other words, whether there are any legal provisions that actually prohibit?
automated driving. Smith says that some states such as Texas and some companies such as Uber
have embraced the uncertainty about the legality of driverless vehicles because it gives them room to
operate. They hope that by acting boldly they'll set precedents that eventually will solidify into law.
Other companies, especially large, well-established ones, will be more cautious and try to use
their market power and political clout to shape policy in ways that will be friendly to them.
Each state will have to find its own solution, whether through legislation or executive order.
All these variety of approaches to addressing questions of legality, the keys that none of them are actually providing the clarity that their proponents purport.
So even in states that have explicitly addressed automated driving, there are still going to be questions about who the driver really is for any specific technology or application.
My state, New York, has an unusual law that requires drivers to keep one hand on the wheel, which would be a big obstacle.
if there were no wheel.
But Smith assures me that this won't be a problem
once we have the completely autonomous vehicles
that I would need to ride alone without a license.
In a truly self-driving vehicle,
you won't have any obligation to keep your hands anywhere
because you just won't be the driver,
and so it won't apply to you.
The question will be who will be the driver, if anybody,
and will they be sitting in a call center 100 miles away
with their hand on some imaginary wheel
or on a computer mouse
to comply nominally with that law?
The New York State Senate just passed a law that will allow driverless cars on public highways until April 2018,
but only for tests and demonstrations, and only under police supervision and with a licensed driver on board.
Smith says that the federal government, too, has recently updated its auto vehicle policies in ways that have given some guidance to states
and potentially provided a path for the feds to exert greater control.
Last year, for instance, the U.S. National Highway Traffic Safety Administration wrote in a letter to Google
that an artificial intelligence operating a car could be considered the same as a human driver from a legal
perspective. But that statement also included caveats about the need to demonstrate that a fully
autonomous vehicle could meet the same safety standards applied to existing cars, as well as the
need to rewrite laws to remove requirements for current car fixtures, such as dashboard alerts,
brake pedals, and steering wheels. This regulatory roadmap is complicated by the fact that autonomous
vehicles still fall into a gray area where it's unclear who has jurisdiction.
Historically, the federal government has regulated new motor vehicle design, while states have
regulated at least non-commercial drivers and driving. And so here we have an instance where the
vehicle is becoming the driver, and that tends to blur some of the lines between the federal
role and the state role. Things could also get complicated on the insurance side, depending on
whether people own cars or rent rides from companies. If Uber and Lyft own all the cars,
then they're liable if something goes wrong. But if a person owns the car, responsibility for
at least some accidents could be debated. For instance, if a car's firmware isn't updated and an
accident occurs, is that the user's fault or the manufacturer's fault for Reading subpar software
in the first place? As Mark Allenbeck of the Washington State Transportation Center says,
figuring all of this out isn't going to be simple. The U.S. has this real mix of land uses,
land types. So we're a great laboratory in how those kinds of policies will take place.
So I think you'll see a lot of experimentation, a lot of fighting from the different business
interest. It's easy to imagine what we'd gain from a driverless society, even a
Aside from my not having to hide my lack of a license,
Lyft President John Zimmer has written about driverless cities
with narrower streets, wider walkways, and parks replacing parking spaces,
and Lyfts Tagger-Metisan envisions a more peaceful near future,
where driverless vehicles are seamlessly integrated into our existing reliance on ride-sharing apps
rather than primarily owned by individuals.
The app would automatically assign you a driverless vehicle whenever the company was confident
that one could deliver you to your destination.
So as a passenger in New York, you're going to open up your app,
you're going to want to go from point A to point B.
And if that is a route that we can service by way of an autonomous vehicle,
then you will see on your mobile app very similar today in terms of matching
that you were matched with an autonomous vehicle.
Mattson acknowledges that there are some basic questions that remain unanswered,
such as how to ensure that the driverless vehicle unlocks only for the intended passenger
and doesn't start driving until its riders are ready.
But for Mattsin, the autonomous future is a utopian one,
where we all have less road rage and spend less time in traffic,
in addition to being able to do without driver's ed.
We are living in this world of just insanity in terms of traffic.
I do imagine how could this all change?
What if there was more orchestration in terms of intelligent cars?
What if we had less cars on the street?
We wouldn't have all of these people spending all of this time looking for parking,
let alone even driving.
And that's where I get excited about how people can kind of take time back
and really kind of focus on themselves.
In Manhattan, where I'm wrestling with my driving decision,
driverless cars have a lot to contend with.
Not only all the cars driven by us error-prone humans, but pedestrians, construction, bad weather, and police directing traffic.
In some ways, that makes this the worst driverless environment possible.
But as Halenbeck says, it's also the one with the greatest demand.
If you're in a place where taxis are convenient and parking is a pain, then the shared use of an automated vehicle has a really good incentive and has a really good business model.
And therefore, that's going to happen because the costs will drive people to take advantage of them.
But that might not be the case outside of the city where the population density declines.
The, I want to drive to Boston and back, or I want to drive from Ithaca to downtown Manhattan,
that's not a very good business market for an Uber or a Google.
In those cases, you're going to want to own that car.
So you're going to buy the Tesla, you're going to buy the Mercedes-Benz.
Better yet, 10 years from now, you're going to buy a sleeper car,
and the sleeper car is going to let you sleep all the way from Ithaca into Manhattan.
Put that way, it all sounds pretty pleasant.
Hallenbeck, though, sees some potential for a dystopic outcome in which advances in technology can't conquer the opposing forces slowing us down.
For one thing, even when fully autonomous vehicles begin to appear, they won't have the roads to themselves.
They'll still have to contend with their less self-sufficient predecessors.
Helen Beck says the average vehicle fleet turnover takes 12 to 15 years, which means that even if we could introduce completely autonomous vehicles today,
we'd still have to retrofit the old ones or face a long path to implementing an entirely autonomous fleet.
Furthermore, making travel easier changes our incentive to travel.
The utopian view runs smack into behavioral issues.
The behavior side is going to limit how much magic congestion relief you get,
either because we travel more or because we travel in more concentrated spaces.
If automated cars really make travel easier,
what's going to happen is the amount of travel we do grows.
In addition to the same people taking more trips,
there might be more people on the roads,
including the young, the elderly, the disabled, and for that matter, me.
People who aren't licensed to drive but will no longer need to be.
Worse, movement could come to a halt thanks to a new public menace,
the autonomous vehicle traffic troll.
If all cars talk to each other and they all have automation,
they can drive a lot closer together.
And therefore, more cars fit through a given amount of space and a given amount of time,
and that's congestion relief.
Yay.
But as soon as you make a pedestrian cross that road,
pedestrians can't go 25, 30 miles an hour,
the cars have to stop for the pedestrians.
Then you add 13-year-old boys to the process
where 13-year-old boy loves nothing better than to screw with the world, right?
And so they stand in front of the car and make funny faces at it and laugh
because what's the car going to do?
The car's stuck.
It knows it's not allowed to run over the 13-year-old boy.
As a former 13-year-old boy, this is disturbingly plausible.
The result could be an ostensibly safer and easier,
but actually an archic future in which Manhattan won't move.
Can you imagine what happens in Manhattan when New York
realizes that cars can't run them over, they'll just walk across Fifth Avenue. Right now,
you walk across Fifth Avenue, you die. But if the cars are designed to avoid hitting you,
I mean, wow, that talk about power shift to the pedestrian. I mean, I always like to say that
congestion will not go away until we have the Star Trek mass transporter. And then at 8 o'clock
in the morning, there will be a line outside the Star Trek mass transporter so they can beam you
into your office at 758 because who wants to get there any earlier?
The rise of driverless vehicles is bound to have unintended consequences off the road, too.
Even Matissan acknowledges just how jarring that first driverless ride is going to be.
The last hundred or so years, vehicles have been built around the driver.
And the minute you pull that driver out, you go from a very familiar environment to a very foreign one.
Cars in American culture are so tightly linked that there's bound to be some blowback to surrendering control, if not a full-on existential crisis.
John Heitman is a history professor at the University of Dayton and the author of the automobile and
American life. He tells me that his cars came to prominence more than a century ago,
it didn't take long for them to turn into status symbols, luxury items owned by the elite.
That could happen with driverless vehicles, too, until, like personal computers and smartphones,
they become affordable for everyone. What you drove said something about you and your family,
and it's in an America that probably more than any other country in the world is defined by its mobility.
Last year, a Tesla driver was killed on a Florida highway with the car's autopilot engaged,
which touched off fears that the cars would be recalled.
But although the software failed to prevent that collision,
Tesla was absolved of blame and even commended for cutting crashes by 40%.
If vehicular history repeats itself,
driverless vehicles might be demonized initially,
with every mishap like that one seized on
as a harbinger of dangerous technology run amok.
But resistance will likely be overcome quickly.
The same rural folks who would embrace the Model T in that decade before the Model T,
say leading up to 1910 or so,
This was the label they used was the devil wagon.
And they resisted the motor vehicle because you had urban types who would be driving out on countryside tours, would pick their plums and their apples.
You get these old guys like myself who grew up in the 1950s at the golden age, you know, 50s and 60s.
And they might be quite resistant to a fully self-driving vehicle.
But I think their children and their grandchildren will not be.
But it's tough to project what we might lose in the automation process, perhaps without even realizing it.
As Hightman notes, operating an automobile is a manifestation of freedom.
When you drive, you make decisions.
You are the master of your life in many ways.
You can go in any one of a number of directions.
What do we lose in terms of this autonomy if we take the steering wheel away from us?
How is your sense of freedom compromised?
You got to remember your sense of freedom is an illusion in many ways.
But it is a sense of freedom in a world where much of our daily routine is being dictated to us.
Even the stories we see on our screens would have to change to reflect reality.
In a world without human-driven cars, Hollywood would have to find clever alternatives to the chase and crash scenes that dominate action movies today.
If we have self-driving cars.
How are you going to drive down the steps of a European city like Born in a mini if you've got all those features attached and you don't even have a steering wheel anymore?
Of course, there's always something like the car hacking in the fate of the furious.
I want every car in a two-mile radius now.
There's over a thousand of them.
Hack them all.
Or maybe tomorrow's action heroes can kick through windshields and scramble across autonomous car roofs like Tom Cruise and Minority Report.
You can't run, Joe.
Everybody runs.
Even darker, the...
The creative director of the kid-friendly Pixar series Cars recently suggested that those cute cars in the movies were originally autonomous cars that rebelled and eradicated their owners.
Adolescent rule breaking, too, could get even more complicated.
You're a teenager. You don't have a license, but your parents give you access to a driverless car.
What happens if you want to take your date to a shady place and make out?
What's going to happen there?
Among other things, your parents probably will know the route, or they can access the route afterwards.
We're all going to have to get good at clearing our car's history, the way some of us already have with our web browsers or GPS systems.
As Hightman says, technology moves very fast, but our understanding of what that movement does to us lags considerably behind.
It wouldn't be surprising if the stigma attached to non-drivers today.
Yes, I've seen some raised eyebrows and quizzical looks when I've admitted to not having a license,
is one day attached to people who do insist on driving themselves, thereby making the rest of us less safe inside our computerized cocoons.
But I can't forecast the course of human history.
I can only alter my own, and it's almost time for me to make my decision about driving.
As I talked to each of these experts on driverless vehicles, I asked them what they thought I should do,
starting with mobilized and galves.
I'll put it this way.
Like, I have a 10-year-old and a 4-year-old.
I think the 10-year-old will get a license.
I think the 4-year-old, there's potential that he'll never see the need to get a driver's license.
The 4-year-old will be eligible to drive in 12 years.
I've waited 14 since I became eligible for a license, so what's 12 more?
Liss Tackert Mattson, too, says I can wait for driverless vehicles to make a driverless vehicles to make.
driving defunct. To you, I would say hold tight, continue to trust the network and the kind of
availability and freedom that we provide. And over time, you will start to see this introduction
of autonomous vehicles. And Smith, the legal expert, concurs. For your purposes, you are likely to be
traveling at some point in truly driverless vehicles that are the Manhattan equivalent of the taxi
or the driverless Uber. And so no, you won't, as a legal matter, need to have a driver's license.
And as a practical matter, you won't either.
Even Mark Hallenbeck, who laid out the ways in which the rollout of autonomous vehicles might not go so smoothly,
says I should keep procrastinating.
If you don't have a driver's license, I wouldn't bother to go get it.
If you survive without it, your life will get better.
As enlightening as it is to talk to people at the forefront of the driverless car revolution,
though it might also be useful to step outside of that bubble
and consult a regular motorist who can make a case for driving the old-fashioned way.
As it happens, I have a connection to someone who once found herself in the same situation I'm facing now.
I'm Joni Cunningham, and I didn't get my license until I was 30 years old.
As he might have inferred from her accent, Johnny grew up in the Bronx, one of 12 siblings
and the daughter of immigrant parents.
Like me, she either walked everywhere or took public transportation, and when she got
older, she dated guys who had licenses.
But when she was 24, Joney decided to take driving lessons.
She'd gotten married, had a son, one of our podcast producers, Jim, and become a nurse,
getting rides to and from work from her husband and a coworker.
Driving was suddenly something that seemed useful.
I signed up for three or four driving lessons.
Then I went for the rides with the guy to take the lessons, and he pulled over into a little side street.
It wasn't a great neighborhood either on the second lesson.
And I was doing okay.
And he wanted me to parallel park and a parallel parked.
And he said, oh, you're doing so good.
Can I give you a hug?
And I said, uh, no.
The creepy instructor's actions brought an end to Joni's driving lessons.
But sometime later, her life changed again.
She had a second kid and moved to the suburbs.
By this time she was 29.
I decided I had to get my license.
I really had a long walk to anything, and I was confined.
I was working nights.
While Joni was practicing driving her husband's huge caprice,
luck intervened and made it possible for her to buy a more manageable car.
She split a $100 ticket for a church raffle,
and a few weeks later, the church called to say that she'd won $10,000.
Joni used the winnings to buy a used 1983 Toyota Tersell.
She and her husband drove the caprice to pick it up,
and then she took the Tersell home herself.
Its maiden voyage was memorable.
The two kids were with me.
I'd pick up the car.
I go to merge onto the highway.
I'm sitting there a good few minutes because I was afraid to merge.
I was just scared out of my mind.
I couldn't adjust the mirrors.
The window was rolled down and it was freezing.
And Jim is flying in the backseat.
I'm freezing.
I'm freezing.
Please roll up the window.
Roll up the window.
I said, I can't.
I can't roll up the window.
I was paralyzed with fear.
Finally, I got the nerve to merge,
but I didn't get the nerve to roll up the window.
It was supposed to be a 20-minute drive home.
it turned out probably to be a 40-minute drive home. With the window down, I was just so afraid.
It was funny. Once she made it home, she was afraid to drive again, but she forced herself,
going one block farther from home on every trip. Eventually, she was confident enough to drive
to the nearest mall and White Plains, a small city in Westchester. It was about 15 minutes away,
but that experience, too, proved traumatic. I was taking the local roads. I wouldn't dare
get it on the highway again. I get there, I'm, you know, really afraid. And I go to park,
and I really didn't have any experience with parallel parking.
The two kids are with me again.
They go to parallel park over by the mall, and I couldn't do it.
I was trying, and the car was tiny, but I was really trying my best,
and I kept going in and out, in and out.
So some man comes over, so he's standing there, and I thought, oh, thank God, someone's here.
That's someone's going to help me park the car.
So I'm looking at him, and I'm pulling in and I'm pulling out.
And he said, wanted you get your license out of a Cracker Jack box?
That was it.
By then, I was already ready to cry because I couldn't park,
but I started crying and I peeled out of the spot and just left.
Never went shopping.
Never went to the mall.
The kids were both crying.
It was a disaster.
Hearing this, I'm about ready to make up my mind about the whole driving thing.
Up until now, I've only been considering the inconvenience of learning to drive.
But what if I'm afraid to get on the highway?
And even if I'm not, maybe I should be.
Driving is dangerous.
The less I drive, the less likely I am to get killed in a car crash.
Joni says, though, that after the initial terror subsided, driving became routine.
She even believes that starting late made her a better driver.
I've never had an accident.
I've never had a speeding ticket.
I've had two parking tickets.
Definitely better, mature.
I think you take a lot of chances when you're young.
And she also says that driving has really enriched her life.
Getting my license definitely gave me more independence and freedom.
And I was able to take different nursing jobs at
became available. Plus, I was able to drive the kids to them from school, all the activities. So I can't
even imagine if I hadn't driven. Even though I did get my license late, I'm so happy I did that.
Naturally, I have to ask her for advice also. And she says something different from everyone else.
Ben, definitely go for your license right away and get on with this because it opens doors.
You have so much more freedom. You're not dependent on people. It's just a whole different life.
It's just unbelievable. Don't wait.
I feel bad about disobeying Joni.
But if I do decide to wait, how long a wait will I be looking at?
For an autonomous vehicle, ETA, let's go back to Dan Gals.
For an urban environment, fully autonomous, 2020, 2021 is when we think you'll start to see vehicles
with reasonably priced equipment on them, normal looking vehicles.
We think that it will be a two, three year period of data generation in controlled
environments in order to get to the point where you can really prove that the technology
is working, then these people that are the safety monitors that sit in the driver seat cannot be there.
And I think you're in a position 2020, 2025, where you will start to be able to hail a fleet service
that can pick you up with nobody in the driver's seat. By 2030, you could start to see these vehicles
really available at a reasonable cost. That doesn't sound so bad. I think I've reached a verdict. I hate to
let Joni, not to mention my own mom, down. It's possible that someday I'll find myself in the same situation
She did and make the same decision.
Maybe I'll move to the suburbs, have kids, and need to take turns driving the minivan.
Life is unpredictable, and I reserve the right to change my mind.
But for now, and for the foreseeable future, my non-driver's license state ID will suffice.
The freedom and independence Joni needed a Toyota Tercell to get might be possible for me to obtain
without a license if I wait a little longer.
Maybe my generation's slightly creepy computers can replace her generation's extremely creepy
driving instructor.
So I'm going to stick with what's worked for me so far and keep kicking the can down the road,
a road that I hope will soon be full of cars with no steering wheels.
It's hard to say how many Americans are with me on this.
There are some surveys that report that most people are ready for driverless vehicles
and others that say that they're terrified to try them,
with individual attitudes dictated in part by one's awareness of autonomous technology
and experience with driver assistance systems.
Maybe some of you are a little less wary than you were when this episode started.
One thing this inquiry is made clear, as John Heitman says,
is that the arrival of autonomous vehicles is going to be an adjustment for drivers and non-drivers alike.
We live with, it's such a hurry-up life.
I think back to the great Chuck Berry, one of his great songs,
No particular place to go.
You know, it's driving with no particular place to go.
You know, it's driving with his baby, no particular place to go.
That's no longer a part of our culture to any real extent.
cruising is very much something now that's about lost in the midst of time.
And of course, these self-driving cars, they always will have a place to go.
You can't leave without giving them information about where you're going.
On the plus side, no particular place to go is also about the singer's struggle to unbuckle his state seatbelt.
And maybe driverless cars will one day be so safe that we won't even have to wear them.
I for one welcome our new robot car overlords.
After all, I'd hate to have to learn how to turn on windshield wipers.
And at this rate, I might never need to.
Since we're nowhere near developing self-producing podcasts,
a honk of the horn to everyone who had a hand in putting this podcast together,
including producers Joe Fuentes, Jim Cunningham, Kyle Crichton, and Zach Cramm.
I'm Ben Lindberg, and you've been listening to a special tech episode of Channel 33.
And if you see me standing by the side of the road, please stop and pick me up.
