San Francisco’s Self-Driving Mess
Speaker 1: And nobody in it.
Lizzie O’Leary: There’s this viral video from San Francisco that made the rounds last spring. It’s an autonomous car getting pulled over by the cops.
Speaker 1: It’s a short little snippet of a robotaxi, like an autonomous taxi, basically from a company called Cruise that is a subsidiary of General Motors.
Lizzie O’Leary: That’s David Zipper. He’s a visiting fellow at Harvard’s Kennedy School, where he writes about urban policy and mobility.
David: This is crazy.
Lizzie O’Leary: The car in the video is a red and white sedan, and it’s stopped on the side of a busy city street at night.
Speaker 1: Apparently because it had some issue with almost a tail light or or a headlight, some minor matter like that. It pulls over to the side of the road and a cop car pulls up behind it. And this is being recorded by a bystander and a police officer comes out to look into the front window and you hear somebody yelling in the background. There’s no driver in there. Nobody in there. Very sorry along those lines. And then the police officer is kind of befuddled. And as he steps back, first, the car just gently drives off. I’m going to have to watch them as officers sort of waiting there. And to be fair, it doesn’t go all that far. It stopped a little bit later. But yeah, it was the autonomous robotaxi was fleeing the cops.
Lizzie O’Leary: On the one hand, this video is undeniably funny. But on the other, it’s a little concerning that this car was not exactly operating like it’s supposed to. And the cop really had no idea what to do. David, who has been studying autonomous vehicles, especially taxis, says there have been other more worrisome incidents in San Francisco.
Speaker 1: Late one night, again, it was a cruise robotaxi that, according to the city, had halted in traveling alongside a garbage truck that was doing collection. And there was a three alarm fire at the time happening in San Francisco. So a fire engine with its siren blaring was on the way to that fire. And the cruise vehicle just froze and didn’t recognize that it was in the way of the fire engine in a way that a human driver obviously would have. And that fire engine was delayed. It was actually halted there for a bit. And until the basically the garbage trucks staff was able to come out and move their own vehicle. But this is actually something that the San Francisco Fire Department has said, like we’re really worried about response times being impacted by issues involving robo taxis in our city.
Lizzie O’Leary: Today on the show, David says that these incidents should be a major warning sign about autonomous vehicles.
Speaker 1: But I wouldn’t just say put this on the self-driving car companies. I would put it on a failure of of regulators to basically ensure that the transportation network is is functioning and that the most basic needs of city residents in terms of security, let alone, you know, an efficient transportation network, are being protected as these these vehicles are being developed.
Lizzie O’Leary: I’m Lizzie O’Leary and you’re listening to what next, TBD, a show about technology, power and how the future will be determined. Stick with us.
Lizzie O’Leary: It makes sense that San Francisco would be a testing ground for the best and worst of autonomous vehicles or EVs, as they’re sometimes called. After all, it’s right next door to Silicon Valley, where companies have been working on self-driving cars for years. But David says there’s another reason why the city is attractive to RV companies.
Speaker 1: San Francisco does not regulate or permit or have really any involvement with autonomous vehicles in in the city at all. That’s all with the state. So what’s happened is you have literally dozens of companies that have applied to the state to do autonomous vehicle testing. They may be small companies from abroad, from China or something, doing small level testing somewhere in California. But let’s be honest, it’s usually in the Bay Area to Tesla, to various companies like Argo, AI, which just folded a few weeks ago, and Cruise and Waymo, which is part of Alphabet, the Google company.
Speaker 1: And the vehicles you see the most of in San Francisco are really Cruz and Waymo, which are these robo taxi services. They’re permitted to actually bring passengers around, not really testing as much as they’re like a paid Uber substitute, if you will. And those are the ones that you’ll see pretty frequently in certain San Francisco neighborhoods, sometimes without a driver at all, picking somebody up, dropping them off. And every now and then, things go very wrong.
Lizzie O’Leary: We’re going to get deep into the regulatory stuff in a little bit. But. How many of these things are out there? Like, if you take a look at a typical San Francisco street, are you going to see one?
Speaker 1: Great question. Not only do I not know the answer to your question, I don’t think the city knows the answer to your question, because these are statewide permits. Right. So it’s really up to Waymo where they go in the state. This is something that San Francisco, as well as Los Angeles and San Diego saw coming as a problem, like please don’t issue entire statewide permits. Please control where they can be deployed. So we’re not we’re able to manage things a little bit. But what’s called the CPC, the California Public Utilities Commission rejected that request. And so as a result, not only do I not know how many vehicles are in San Francisco.
Lizzie O’Leary: Nobody knows what’s the purpose? Are they are they all taxis?
Speaker 1: That’s what Cruz and Waymo are providing robo taxi service. There are lots of other potential use cases, including potentially eventually being able to buy your own self-driving car, which you cannot do even if Tesla implies otherwise. So that’s marketing. But the idea is that once you ultimately develop the technology, let’s say Cruise solves autonomy and is able to be deployed perfectly anywhere. You could imagine that software hardware kit potentially being deployed into a vehicle that somebody could purchase. But we’re nowhere near that now.
Lizzie O’Leary: Which is pretty clear from some of the incidents I’m using and otherwise that have cropped up in San Francisco.
Speaker 1: Well, there are a lot and we know about them, by the way, not because they’re all necessarily reported, but because somebody tweeted about it or put it on Instagram or perhaps called 911. Most of the stuff that’s been identified by the city and in the public has been involved crews. There have been vehicles reported driving on the sidewalk. Not good. There was a cruise vehicle that froze on the streetcar line for the N line in San Francisco, which held up to over 140 passengers for about 7 minutes until Christmas, able to send somebody to come and physically move the car.
Speaker 1: Waymo has all has had fewer incidents that have been been identified. But one is sort of comical where there was just like a parade of every few minutes, another Waymo Robotaxi, going into this dead end street on 15th Avenue where there’s nowhere to go, just go and then quietly turn around and annoy the hell out of the neighbors.
Lizzie O’Leary: There are some days where it can be up to 50. I mean, it’s literally every 5 minutes. And, you know, I’m. We’re all working from home, and so are we. This is what we hear. Yeah, this is what we hear. For they all kept kind of getting stuck down there, like doing 12 point turns.
Speaker 1: No, exactly. And I mean, some of this stuff really is kind of amusing until you start digging into it with them. It’s like a huge cluster of of like over a dozen cruise vehicles that mucked up another street in San Francisco. And somebody commented, which I thought was funny on Reddit, like, Oh, no, they’re plotting. Like, it really runs the gamut. But there’s the suffice to say there have been regular blockages of public roads and sometimes things like I mentioned, driving on the sidewalk, they’re just flat out illegal and really dangerous. So, you know, joking aside, this is really no way, in my view, to run a transportation network.
Lizzie O’Leary: Well, that’s kind of where I wanted to go next, which is, yes, this is in parts hilarious. But as you mentioned, there are concerns and some of it feels dangerous. Like should people be worried?
Speaker 1: I think so, because importantly, we’re in the very early days of all of this. Right. San Francisco is likely to get more AVI vehicles as these companies ramp up. Now, Cruz is applying to the state and to the federal government to deploy a new vehicle called the Origin, which one would imagine is going to lead to still more of these vehicles being deployed on city streets. I think we should be worried because San Francisco itself has no real power to handle any of this. And it’s not clear that the state is is able to to to step in either. These AV companies want to expand really quickly now and they’re going into other cities around the west and southwest. So more people are going to be experiencing some of the stuff that’s been happening in San Francisco.
Speaker 1: I do think that that we should be concerned because if you scale autonomous vehicles in a given location and you get some pretty bad effects, even if the technology works perfectly, you’re basically making it easier to take a car trip, right? And if something gets easier, we do more of it. People find new reasons to take car trips, and in most cities in the United States, the residents and leaders want fewer car trips. They want to actually move in exactly the opposite direction. So I worry for climate reasons. For safety reasons and just for urban quality of life reasons. I don’t really know that even if the technology works, which it doesn’t work flawlessly now, but even if it does ultimately do that, it’s just not clear to me how this actually enhances urban life.
Lizzie O’Leary: How do the companies respond when one of their cars, you know, gets pulled over by the cops or sits on the sidewalk or blocks public transportation?
Speaker 1: So they have what’s called a critical response line, which the city can contact in an emergency, and then they will deploy someone to address the vehicle if it has no driver in it. However, that doesn’t always work perfectly. I give examples like like San Francisco officials said that there was one instance where they called four times Cruise’s critical response time in 6 minutes. None of those calls were picked up. And when I asked Cruz for comment for the story, they said that their goal is to be responsive within 10 minutes when an emergency is reported. But they didn’t answer a question that I asked of how often they actually achieve that threshold. And by the way, none of this is reported anywhere like this seems like a really clear thing that somebody should be forcing these companies to report. Nobody’s doing it.
Lizzie O’Leary: So there’s no central data repository.
Speaker 1: Not at any level of government. And and it’s all. And so it doesn’t exist anywhere.
Lizzie O’Leary: Well, let’s talk a little bit about I guess I would say the the opposing forces. Right. Pushing for these vehicles and then regulating or supposedly regulating them, I guess, first off, who who wants these cars to be driving around San Francisco? Presumably, you know, Cruz and Waymo do. But is this a service the public demanded? Is this venture capital companies saying we want some return on our investment or is this at the behest of shareholders?
Speaker 1: Well, it’s certainly that the pressure from venture capital investment dollars, this is an incredibly expensive product to develop autonomous driving systems. And some of these companies have raised literally billions of dollars. So they have to show some kind of return on investment and some sort of evidence that not only can the technology function in on public streets, but that people will be willing to pay for it in some capacity. And that’s what some people are doing now with Robotaxi services and paying the fee for it.
Speaker 1: But you ask a question that I frankly struggle with. I keep asking companies this myself and investors this myself and Avi employees. I’m saying, what exactly is the purpose of this? When we scale this, how how does it make society better off and who’s asking for it? And it’s an open question to some degree. I worry that the whole premise of autonomous vehicles is one based off of the idea of or the interests of people who live in in suburban or ex-urban locations and are just tired of driving and want to not have to worry about that anymore. That’s a vision that might exist. But I frankly think that’s why we could talk about it. But I don’t know that that we want to actually make it easier to live far away from a central city and basically take up highway space on the way in because you’re in a driverless car because that’s terrible for the environment. And I’m not sure what cities get out of that.
Lizzie O’Leary: Is there an argument that autonomous vehicles are helpful for people with disabilities?
Speaker 1: There is. And that’s those people who are vision impaired sometimes say that this will be liberating for them. And I guess what I scratch my head over a little bit is, you know, if if if that’s the issue, if you’re mobility impaired, you still can call, you know, a ride hail vehicle or a taxi. Now, it’s just the difference is you no longer have the driver there. So I know I’ve written about this before. I just scratched my head a bit trying to figure out what the value add is for a self-driving car vs v a human driven car. You could maybe argue safety, which these companies do sometimes, but that’s really speculative because autonomous vehicle systems will make certain they’ll make fewer certain kinds of mistakes to humans while they’ll never drive drunk. But they’re also going to make mistakes the humans would not like maybe driving away from a police officer when they’re not necessary. Even it may look like they’re trying to flee. They really weren’t. It was just bad programming.
Lizzie O’Leary: When we come back, Who exactly is overseeing all of this? One of the most confusing questions about the self-driving cars on the streets of San Francisco is who is supposed to be regulating them? David says it’s a bit of an alphabet soup.
Speaker 1: So at the federal level, there is nits, which stands for the National Highway Traffic Safety Administration, and they’re the ones that are really responsible for ensuring the safety of vehicles on public roads in the United States. So AV companies sort of ask for permission to deviate from vehicle safety rules that are really set up for driver driven cars. And when things go wrong with Avs and it’s they can basically to issue a stop order and then they have to halt deployments.
Speaker 1: But California has basically its own state oversight, which by the way, kind of annoys some of these companies. They wish the state wouldn’t be so involved. But there is the Department of Motor Vehicles and the Public Utilities Commission that regulate the design of the vehicle on California roads and then the passenger service, respectively. It’s quite confusing, but it’s really at the state level that that these that this oversight takes place. And cities and counties really play no role at all, which is part of why you end up with the weird situations you do in San Francisco.
Lizzie O’Leary: And that led to a particularly strange moment this fall when two San Francisco agencies basically went over the state’s head. Writing a letter to the National Highway Traffic Safety Administration, or Nysa with concerns about AVS. It seemed to me like they were saying, Hey, we need some help, but we, the city, are not in charge of these things that are driving around on our streets.
Speaker 1: You’re exactly right. It’s extremely unusual for a city transportation agency to put together a 39 page letter to net side of the federal government pointing out problems with Robotaxi deployments in their city. Because you’re right, without saying so. San Francisco basically did an end run around the state, raising these issues to the federal authority. And I think it’s not a crazy inference to suggest that the intent was to perhaps put some pressure on the state to take the city’s concerns more seriously. That’s to be clear. That’s my own read. No one’s confirmed that to me. But I will say that you will not find many documents of this kind that are directed at the local level, straight to the feds without really seeking the the state people very clearly.
Lizzie O’Leary: Maybe this is a leap on my part. But as someone who covered the financial crisis, it reminds me a little bit of the regulatory arbitrage that was going on between big banks and and investment banks and financial companies trying to figure out like who was maybe the lightest regulatory touch in terms of what they the company wanted. Is is that do you think, a fair analogy or an unfair one?
Speaker 1: Yeah, I think you’re on to something. And there’s a track record of this in transportation, too. You know, when Ride Hail first emerged, it was unclear who was going to regulate it, the cities or the states themselves. And Uber and Lyft, I think, very, very quickly realized they did better off if the states, rather than cities, were managing what they were doing because they would get less pushback and less demands for data and less concerns about traffic congestion and taking away transit riders and stuff like that.
Speaker 1: So almost everywhere the ride hailing managed now at the state level, not the city level. And that has created huge headaches for city transportation departments.
Speaker 1: And frankly, I think that that experience, which is only like ten years ago, colors a lot of what city transportation officials now or city residents look at now when they see these thinking, oh, gosh, are we is history about to repeat itself and we’re going to end up sort of being left it left out as like our own streets get mucked up by a service that we can’t control.
Lizzie O’Leary: You alluded to this, but San Francisco is sort of the testing ground for this stuff, but it is obviously not the end game for autonomous vehicles. Where are they going next and are they going to places that might have less robust regulatory environments than California? After all, California, even if the state is regulating the stuff, is a lot tougher than many other states.
Speaker 1: That’s right. Yeah. So about two months ago, Argo AI ceased operations. And this is a company.
Lizzie O’Leary: Another one of these companies.
Speaker 1: Yeah, that may be coming. That brings literally billions of dollars in investment, especially from Ford and VW. And that was a very well respected company by a lot of folks. And that closure really sent shockwaves through the RV industry and made investors get nervous about the ability to actually. Developed this technology in a sort of viable way, and it put a lot of pressure on the remaining EV companies to deploy in more places to show that, hey, we can do it. Our technology product works.
Speaker 1: So you’ve seen sort of a quickening of the pace, if you will, of announcements of EV companies expanding into new cities, and that includes Cruise and Waymo, as well as a couple of others like Motional that are expanding into cities like Las Vegas, Los Angeles, Phoenix, Austin, all of them warm weather cities. With light regulatory oversight at the state level, except for California, which is its own weird beast, and notably no snow. Because one of the little secrets of AV technology is it doesn’t work in bad winter weather. So if you live in Cleveland, you’re not going to be seeing Avs for quite a while.
Lizzie O’Leary: To push back. Is there a case to be made that in, I don’t know, a a western city, some place with less traffic or less congestion than a San Francisco that you might not see some of the issues that you’ve seen in San Francisco?
Speaker 1: I think maybe. I think that’s possible. And, you know, I’m not basically out there sort of like rooting for all of these AV companies to fail. But what I do think is that there are real risks. If nothing else, the experience of San Francisco shows things can go wrong in terms of individual incidents that are public safety risks, but also like take a step back, like like we may end up with a lot more congestion, a lot other sort of systemic problems that get in the way of traffic planning. And this is why I take such a regular regulation. Heavy perspective on the topic is is whether or not that small city you’re describing has fewer incidents of the kind that San Francisco is enduring.
Speaker 1: Still, those local officials, as well as the state government, should be looking carefully to make sure there’s transparency about what is going wrong, whether it’s a lot or a little bit, and that there’s clear data being collected about about how the net these autonomous vehicles are affecting the transportation system, because that is just not happening in California, which, yes, it does have a stronger regulatory perspective than Arizona or Nevada or Texas, but it’s not a very effective regulatory structure at all. And I frankly think it’s in pretty urgent need of a reboot.
Lizzie O’Leary: You know, for so long, Silicon Valley companies and car companies have been working to make driverless cars. And frankly, reporters have covered them kind of breathlessly. And now that they’re here, I wonder, what do you think we’ve learned?
Speaker 1: We’ve learned how to burn up a lot of money. I think some people have learned about the hype cycle and how, you know, you don’t like it.
Speaker 1: If you are really excited about a new technology forecasting the that the date at which we’re all going to be using that technology, do you do that at your own peril? Because there’s a lot of people with egg on their face, with their sort of beautifully crafted graphs about AB uptake that they issued in like 2017. There’s one that’s a beautiful graph. Some graphic designers, very skilled, created this for Lyft, showing that by 2022 we’d be gently shifting all of us toward total Robotaxi is everywhere across the country. Beautiful graph, Utterly wrong.
Speaker 1: But I think what I worry about that we haven’t learned as a society is that technology is not a panacea. That goes for for things like Hyperloop, which is also failing to meet expectations. And it goes for a number of other technologies as well. I think that there is a lesson here about being realistic about sort of mundane mobility solutions being better than the new crazy technological ones. And to recognize, too, that autonomous or even, dare I say it, electric cars are not going to remove some of the fundamental problems of cars, period. In our cities, dense cities like San Francisco don’t function well when they’re overrun with cars. I don’t care if you make it autonomous. I don’t care if you make it electric. And that’s a lesson that I think we’re starting maybe to learn. But I don’t think we’re there quite yet.
Lizzie O’Leary: David Zipper, Thank you so much for talking with me.
Speaker 1: It’s been a pleasure.
Lizzie O’Leary: David Zipper is a visiting fellow at the Harvard Kennedy School’s Taubman Center for State and Local Government. You can find his work on Slate and at David Zipper dot com. And that is it for our show today. What next? TBD is produced by Evan Campbell. Our show is edited by Tori Bosch. Joanne Levine is the executive producer for What next? Alicia montgomery is vice president of Audio for Slate. TBD is part of the larger What Next Family. And we’re also part of Future Tense, a partnership with Slate, Arizona State University and New America. And hey, if you like what you hear, I have a little request for you. Become a Slate Plus member. Just head on over to Slate.com. Slash, what next? To sign up. It makes a lovely holiday gift.
Lizzie O’Leary: All right. We will be back next week with more episodes and hopefully my voice. I’m Lizzie O’Leary. Thanks so much for listening.
