Ready or not, driverless cars are here (2023)

This transcript was created using speech recognition software. Although it has been checked by a human scribe, it may contain errors. Please review the audio of the episode before citing this recording, and email transcripts@nytimes.com with any questions.

I'm Michael Barbaro of The New York Times. This is the "Daily".

Now, after a closely watched vote, the once Silicon Valley fantasy of self-driving cars 24 hours a day in San Francisco has become a reality. My colleague Cade Metz described the unique challenges of co-existing with autonomous vehicles.

Today is Wednesday, August 23rd.

Cade, we invite you to come to our studio in New York and tell us what's going on in San Francisco. That sounds inefficient, but San Francisco has become the capital of the grand experiment in self-driving cars anyway. I want you to describe what it's like to live in San Francisco in this new reality.

When you walk down the street, you'll find hundreds of these cars driving almost non-stop across most of the city. Once you leave the downtown financial district, you walk a block and you see one of these cars. If this is your first time in San Francisco, or the first time you've been there in years, you won't believe what you're seeing - hundreds of cars driving around with no one in the driver's seat.

These cars aren't hard to spot. They are equipped with various sensors designed to detect everything that is happening in the world around them. These are large, sometimes dumpster-sized, rotary sensors. And you can't-

I saw a mix of a police car and a kid with braces.

I like to use the Ghostbusters car analogy, which is the car from the original Ghostbusters in 1984 with all these things on it.

Well, you've been in these cars. I just wanted you to describe the whole experience of getting into a car and how to get a self-driving car. Because it's clearly no plain old hail system.

Let's recap my most recent ride, which was last Wednesday. I opened the Waymo app, where I could hail a car just like I would an Uber. The difference is that it doesn't take me where I want it to take me. He gave me a seat in an alley a block and a half away where there was less traffic.

Finally my car stopped. I opened the app again and there was a little button on the app that said "Unlock".

As I got into the car, a disembodied voice came through the speakers calling my name. "Hello, Cade."

It also provided me with a safety briefing.

Then the wheel starts spinning on its own. The car starts moving. It was like being in the car with my daughter who just turned 16 and is learning to drive.

Beside!

yes and no. I'm nervous because my daughter is a new driver, but I also have a very serious daughter. She was driving very carefully, stopping a few feet before reaching the intersection. But so are these self-driving cars. They drive fast and are looked after by a very conscientious 16-year-old driver.

[lol]:

Due to the overly cautious nature of the car, it took significantly longer than an Uber when continuing to drive up and down the San Francisco hills.

[music tone]

So, all in all, how do you rate this ride, this weird, driverless, discreet version of a taxi ride?

It got me where I wanted to go, but it took me a relatively long time to get there. Now, on the other side, it's also like at Disney World, where it's a real novelty. This is something you don't necessarily experience every day. As you embark on this journey, your senses are sharpened and you become aware of this strange thing you are doing.

Like many such innovations in the real world, it has been the subject of intense debate, especially in San Francisco, where the cars are often involved in accidents. In some cases, these accidents result in injuries. They crashed into an emergency vehicle. They do things they shouldn't.

That's increasingly a concern for many San Francisco residents, who don't want their city to become a petri dish for cars like these that could cause serious problems in everyday life.

correct. I'm rethinking what you just said. It's like taking a ride to Disney World, but San Francisco isn't Disney World. This is a city with public streets and pedestrians. So it makes me wonder how we got here, is an experiment on the scale you describe even possible. So tell us the story of how we arrived at this moment in San Francisco filled with these wonderful driverless cars.

Scientists and researchers have been trying to build self-driving cars since the 1960s. That goal, largely driven by a Department of Defense-funded lab, is building self-driving cars for the military. It is arguably better to have a vehicle without a driver, a tank.

If hit, it is destroyed.

Exactly. That's roughly the goal. But over time, people started to realize that you can do this for civilians, you can take the same technology and use it to drive regular cars and potentially change our lives, potentially make cars that are safer than Humanity.

Technology never tires. Technology doesn't drink whiskey and drive. Technology-driven has a lot of advantages over humans if you can get technology to the level we need it to be.

When did this start to happen?

It started around 2005. The Defense Advanced Research Projects Agency (DARPA), the research arm of the Department of Defense, hosted a competition. The idea is to ask various research groups to build a self-driving car and then compete against each other to see who is the most capable. Two people noticed this happening.

They happen to be Google founders Larry Page and Sergey Brin. Sebastian Thrun, leader of one of the research groups, tells the story of Larry Page, who showed up at one of the games in disguise—wearing a hat, Sunglasses so you don't get recognized.

What ended up happening was that Page hired Thrun to run a new lab at Google called Google X. The idea is to pursue what Page and his colleagues call moonshots, extremely ambitious technology projects. The first thing they decided to pursue was driverless cars.

Absolute moon.

Absolute moon. Larry Page wants to put one of these things in your driveway so that you can ride every day in a car that can drive itself and take you where you want to go while you're doing other things.

You take a nap and read to your child.

Absolutely. This is something that gets a lot of people excited. But there's an interesting difference of opinion between Thrun and Page that explains the evolution of the technology over the next 10 years. Larry Page, who runs companies like Google, wants to see that happen.

Sebastian Thrun, whom he hired to run the lab and project, realized how difficult it was going to be. He realizes how far they still have to go. And he doesn't necessarily know how to get there.

Why is it so challenging? Why, in Thrun's view, will it be so difficult to realize Larry Page's vision of a driverless car on every road?

It is not difficult to get a car to turn left at an intersection. Turning right is not difficult. Acceleration, braking, these things are not difficult. The hard part is dealing with all the chaos we have every time we get in the car, accidents can happen anytime.

First, you have to have a machine that can fully recognize what's going on around it. Be aware that this is a pedestrian and not a bicycle. Then, once you have a clear picture of what's going on around you, you need to anticipate what's going to happen with everything. Then, once you predict what's going to happen, you have to find a way to deal with it all. This requires a reason. It requires qualities that machines don't have today, and certainly didn't have in the early 2000s.

correct. Driving a car is not just about steering a wheel while driving on the road. It's about reacting to all the things that can go wrong. Dogs running into the street, bicycles suddenly gliding into your path — we as humans learn to respond to them instinctively. Teaching a robot to do this is even harder.

correct.

But despite how difficult it is to teach robots, we know this driverless car technology is moving forward. Please tell us how this happened.

Google founder Larry Page said just go ahead and do it. Thrawn followed suit.

correct. How high do you jump?

So in 2009, Thrun and his colleagues at the new Google X lab set out to build a self-driving car. By 2010, they quietly began testing the technology on public roads in California.

A Secret Driverless Car Project Getting These Driverless Cars On The Road?

That wasn't until the New York Times and my predecessor, John Markoff, heard about the project and let the world know it was happening in the fall of 2010.

What is the response?

Well, this is triggering a technological arms race, not just between dedicated tech companies like Google, but also between new ride-hailing companies Uber and Lyft and traditional car companies that now see their way of life threatened . work. They also join in and start competing.

As companies like Google and Uber start talking publicly about the technology --

They talk about it like it's just around the corner.

If you read the tech press, if you read the mainstream press at the time --

- you guessed next year -

There's a driverless car in your driveway, if you will.

But that never happened.

There is some hype.

Silicon Valley is always full of hype. In this case, the hype is especially strong.

OK So what happens now that all these big companies are pouring money into this unproven technology?

On the face of it, the type of car being built looks perfect. Reporters will be boarded in Mountain View, California. If a mistake is made, they circle the block. But the reality beneath the surface is that manufacturing the technology is extremely difficult, and that has been the case for years, and the hype doesn't reflect reality. But that changed in 2018.

Which is when?

An Uber test vehicle crashed in Arizona. A 49-year-old woman was cycling across a multi-lane road at night. He was shot and died. There was a safety driver behind the wheel who was supposed to take over if something went wrong, but they didn't.

On top of that, someone was killed by a self-driving car, which almost immediately set the industry back. Yes, motorists are involved in accidents every day. People die from these accidents. But if a new technology kills someone, the public and regulators are much more concerned. They start to wonder if this works.

Immediately after the accident, Uber stopped testing those vehicles, not only in Arizona but also in California and Pittsburgh. Toyota is not far behind. Slowly, all of these companies are starting to admit that it's harder than they let on.

Even Google is starting to change the way it does things. Its self-driving car business is being spun off into a new company called Waymo. Its ambitions are starting to shift from the concept of cars to the type of robo-taxi service Uber and Lyft are developing. Instead of having a car in your driveway that can go anywhere, what these companies have decided to do is let us limit self-driving services to specific geographic areas.

Limit variables.

Here's how it works. If you put a ring around San Francisco, you can confine your car to that city, you can create a digital map, a true 3D visual map of the city, and then you can feed that map to the car. Then you run the test again and again in that limited environment.

You limit chaos, you limit unexpected moments. This will get easier, which is what the industry is starting to pursue. As a result, San Francisco, the second most populous city in the United States, is becoming a testing ground for these self-driving cars, with hundreds of vehicles driving around the city.

This has been going on for years, with many companies testing these vehicles to make sure they are safe. But when they try—

-- for motorists and other citizens of the city it is becoming increasingly clear that they are flawed --

—they make mistakes—

— They can clear traffic —

— They can cause accidents —

- They can cause harm.

All of this culminated in public hearings in California earlier this month that will determine the companies' future in the city. The question on the table is, can they operate in cities like Uber or Lyft, without a driver behind the wheel?

But the significance of this decision is even greater. If these companies don't get approval in their flagship city of San Francisco, what happens in other cities across the country?

In a sense, this is a conference about the future of driverless cars.

Absolutely.

We'll be right back.

Well, Cade, set the scene for us at this conference in San Francisco on the future of self-driving cars.

At the meeting, four state regulators will decide the future of two different companies, GM's Cruise and Google's Waymo. San Francisco is a particularly interesting place for this kind of voting. Activism is a part of this city, it has been for decades, and people don't hesitate to voice their views and voice their concerns about the technology.

But at the same time, it is also a technological innovation center for the country and the world. There will be an equally strong sense that this must happen, and it must happen now.

The two parties in San Francisco will meet at the public hearing on Aug. 10.

Regulators will determine its future.

You get a group on Waymo and Cruise, and they immediately start paying for rides.

This includes cyclists --

— People with disabilities, health professionals.

There are all sorts of other reasons why these cars should be allowed on the road right away.

Most importantly, it seems, they wrangled around security issues.

Security is the main argument, but there are others as well.

My name is Sean Durkin. I am the rider of the driverless vehicle. As a gender non-conforming individual, I have experienced drivers pulling me over and refusing to open their cars.

The technology doesn't discriminate like a driver, a taxi driver who can overtake someone on the street.

Then someone stood up and gave classic Silicon Valley reasons.

Technological progress has to happen.

Damn your poodle.

Exactly.

Thank you, members. My name is Michael Smith and I am a technical co-founder of a successful startup in San Francisco. When technology is useful, safe, and actually works, I strongly support it.

Of course, there are just as many arguments on the other side.

Some argue it's not as safe as a human driver.

This will prevent fire trucks and police cars from getting where they need to go.

People also argue, as they often do when new technologies emerge—

This will delete the existing job.

Uber drivers who now make a living driving in cities will be replaced by self-driving cars.

All kinds of people appeared——

— they’re just mad at the tech industry.

This anger has been going on in San Francisco for years --

- Really change the structure of the city as tech workers take over the city.

There is also the cost of the city.

There is also the cost of the city. And some people are just fed up, and it's an outlet for their anger.

correct. Because what better embodiment of technology and the invisible power of its money and everything people might not like than a truly driverless car?

Sometimes it is difficult to define technology and the changes it brings. It's a natural technique in a natural place that creates real chaos in the city.

OK So all these people voted, but in the end, like you said, only four people actually voted. So how did this vote go?

The committee voted 3-1 in favor of Waymo and Cruise to immediately allow paid driverless rides in San Francisco. It shows that these companies, too, can go national with their own ambitions.

In that sense, they opened Pandora's box without guidance. So what happens after this vote?

We get a glimpse of the inside of this Pandora's box almost immediately. Within a week, Tom Cruise had two eye-opening accidents. One of the cruisers misidentified the construction area and drove into wet cement. Talk about a moment you can easily share over and over on social media.

I think so.

Indeed.

Such car accidents have been happening in the city for years now. But in this vote, the technology was more scrutinized, and people started to notice what was happening.

Either right or wrong. correct.

A few days after the cement incident, a passenger in a Cruise vehicle was injured in a fire truck accident in San Francisco.

The next day, the California DMV asked Cruise to cut its fleet of driverless cars in half.

Wow.

So they reduced the number of cars from about 400 to 200.

Well, this is interesting. So the state board voted to allow the technology to roll out, and then a few days, a week or so later, the DMV came back and asked to stop the technology. What do you think about this?

Well, it's been obvious for a long time that these cars have flaws, the question is how much tolerance we as a society have for those flaws. How much chaos, how many accidents, how much injury is acceptable when we get there on the road to this utopia where technology solves everything by itself and makes the world safe?

Well, I think that's a good question. How much tolerance should there be for driverless cars? Because the incident you just mentioned about cars smashing into wet concrete is embarrassing, but does it fundamentally change our understanding of the safety of these cars?

Isn’t this dysfunction a natural process of technological progress? There is a transition period. It's going to be difficult before it's finally widely adopted. Or are self-driving cars just a special case? Does taking a person out of a moving two-ton piece of metal mean our tolerance will be lower? What is your opinion on this? How should we think about it?

Our tolerance will definitely be lower. This is how we think. As humans, we trust in human skill and ingenuity, and trust our own judgment to make the right decision when life is at stake. Do we trust the technology, which in many ways is not as experienced as we are, which does not reason like we do, which does not have the common sense that you and we really depend on and depend on others?

He doesn't have that. It also has a lot of stuff that we don't have. It has more sensors. It can process data faster in some ways, but it doesn't have what makes us human. Do we believe that something so different from us can replace us? This is a difficult question for anyone to answer. That's the question we're trying to answer here.

correct. I think that raises a question that goes back to the beginning of our discussions and the moon that Google co-founder Larry Page introduced. The question is, is this the driverless car we all really need?

This is still an open question. We don't know if we really need it. Even if we were sure of that, it's not clear that the technology would be able to meet that need. We don't yet know if it's safer than a human driver. We don't yet know if it's cheaper than a human driver, and it should be.

What happened in San Francisco was an opportunity to answer that question, to see if there was a need and if technology could meet that need.

When will we get an answer?

[lol]:

I won't let you hold on.

It took 13 years to get this far. It will take us years to truly answer that question.

Well, Cade, thank you very much.

Thanks.

We'll be right back.

Here's what else you need to know today. The Times reports that Ukraine's counteroffensive against Russia may falter because it has placed too many troops in the wrong places. A central goal of the counteroffensive is to cut off Russian supply lines in southern Ukraine, but U.S. military leaders say Ukrainian commanders have spread forces and firepower too far between southern and eastern Ukraine.

Eight candidates have qualified for tonight's first Republican presidential debate in Milwaukee, which front-runner Donald Trump has declined to participate in. Candidates include Florida Governor Ron DeSantis, former Trump Vice President Mike Pence, former governors Chris Christie and Nikki Haley, South Carolina Senators Tim Scott and Conservative businessman Vivek Ramaswamy. The debate will begin at 9:00 p.m. ET.

Today's episode is produced by Rikki Novetsky, Olivia Natt and Luke Vander Ploeg, with assistance from Shannon Lin and Jessica Cheung. It was edited by Devon Taylor with assistance from Michael Benoist, Paige Cowett and Lisa Chow, to original music by Marion Lozano, Rowan Niemisto and Dan Powell, and designed by Chris Wood. Our theme music was composed by Wunderly's Jim Brandenburg and Ben Landsverk.

This is for The Daily. I'm Michael Barbaro. See you tomorrow.