The Self-Driving Dilemma

Hold on to your butts as we take a ride through the self-driving car revolution to see if the future of transportation is a techno-utopia or another way for Big Tech to grab the wheel of our lives.

While some see a world where you can nap through a traffic jam, others see a data-hungry machine poised to replace human workers. In our sixth episode this season, we ride along in a Waymo through the streets of Austin, Texas, to explore how "the law of the newly possible" is struggling to keep up with cars that see the world through lasers and algorithms.

Who gets the ticket when an autonomous car breaks the law? And what happens to rideshare drivers when the "human" is removed from the equation?

Special guests:

• Bryant Walker Smith, professor at the University of South Carolina
• Nassim Parvin, professor at the University of Washington Information School
• Sergio Avedian, contributor to The Rideshare Guy and The Driverless Digest

This episode was produced by Kirk McDaniel. Intro music by The Dead Pens. 

Editorial staff is Ryan Abbott, Sean Duffy and Jamie Ross.

Transcript

(Intro music)
Clip: We're all set for auto control. 
Kelsey Reichmann: Welcome to Sidebar, a podcast by Courthouse News Service. I'm your host, Kelsey Reichmann, coming to you from the nation's capital and one of the most walkable cities in the U.S., Washington, D.C. I'm joined by Sidebar’s Kirk McDaniel, from the much less walkable Austin, Texas. 
Kirk McDaniel: Hey, Kelsey. Everything is bigger in Texas, and all that size means you've got to have a car to get places. For all my life, I've depended on a car to go wherever I want, and it's nice to have that freedom, but nothing makes me feel like I'm wasting my life away more than while I sit in traffic. 
KR: At least on the metro, I can distract myself with a good book, but I definitely don't recommend doing that when you're behind the wheel. Even with top-tier public transportation, walkability and a lot of bike lanes, D.C. is hardly car free. And there's been a lot of talk about adding self-driving cars to our fleet of transportation options. 
KM: Well, get ready for things to get weird. Austin has been at the center of the self-driving car revolution. I'll never forget the first time I saw one of these cars going on a test ride around downtown Austin. It was so eerie to stop at a red light, look across to the car next to me and see no one in the driver's seat. And that was just in 2023. Today, six driverless car companies are looking to provide services here, with two of them already taking folks where they need to go. 
KR: As I hear it, these companies are moving at high speed to be in every city. 
KM: They certainly are. For some, driverless cars foretell of a techno utopia where you can nap in traffic and avoid car accidents. 
KR: Or we're headed for but another trap laid by Big Tech to gain even more control of our lives. 
KM: Maybe by the end of this tour, we'll know if the future is bright behind the wheel of a driverless car. 
KR: So, Kirk, have you ever ridden in a self-driving car? 
KM: Before working on this story? No, but I jumped at the opportunity to try one out. So, on a beautiful April day, I stood outside my home, opened up the Uber app, and called up a Waymo. That's one of the self-driving car companies with services here in Austin. It picked me up and took me on a ride to the Texas Capitol. OK, I think it's finally come to a stop now. OK, car is unlocked. 
Waymo: Hey, Kirk. 
KM: Oh! Hey, Waymo. All right. We have some ethereal sounds happening here in the car. Go ahead and get buckled up. So, what I'm seeing right now, it says, ‘Good afternoon, Kirk,’ on the dash. That's pretty cool. And then it gives me a button to start ride, so we'll go ahead and get started. 
Waymo: Hello from Waymo. As we get going, just give us one minute to cover a few riding tips. All this experience may feel futuristic, but the need to buckle up is the same as always. So, keep your seat belt fastened. 
KR: I wasn't expecting the car to talk to you. 
KM: You heard it, Kelsey. It's futuristic. The car itself is unique. Waymo uses a Jaguar I-Pace SUV outfitted with these camera-looking devices on the roof and near the head and taillights. These devices are actually a combination of cameras, radar and light detection and ranging sensors. These are lasers that can measure distance, which the car uses with its software to make driving decisions. On the inside, there are touch screens on the central and center console that tell me all sorts of information and let me personalize my ride. There's a couple other features in here so I can actually set the temperature. I can tell the car to pull over. I can also tell it to play music. And then there's also a support button here, which was told as we got in the car that if there's any, you know, any issues or anything, that's the moment that I can talk to a person. There's another button here. Let's see what this does. OK. So yeah, this tells us route view, we can look at things about the car, riding tips, camera privacy. Let's just stick to route view. Yeah, I don't know. I keep finding myself wanting to just go back and watch the road as we drive. 
KR: What made watching the screen so interesting? 
KM: It was like a live view of what the car was seeing, rendered in a simple 3D representation. You could see cars around you, pedestrians, cyclists, all as these figures like pieces on a game board. Oh, we're actually passing another Waymo right now. Waymo is very interesting, also show up as like a little, like a little Waymo icon. 
KR: Sounds like you were having fun. Let's hear how the rest of your ride went. 
KM: I'd say we're right now in pretty much the heart of the city, which is Congress Street. It's very beautiful, real long street that runs all the way up to the Texas Capitol. I don't know, this is quite amazing. I mean, this is the culmination of over a hundred years of people thinking about cars, how to make them safer. And here I am, sitting in the back of one about to arrive at my destination. And it's hard not to think that this will be the preferable, you know, mode of transportation at some point. Alright, we're coming up on our destination here. 
Waymo: For your safety, the doors will remain locked when we arrive. Pull the handle twice to exit. The first pull unlocks, the second opens the door. You're here. Please make sure it's clear before exiting. 
KM: Alright. Ride's over. Time to get out. It was fun. Thanks, Waymo. 
KR: Overall, sounds like a positive experience for you. 
KM: It was quite eye opening, to say the least. While I started the ride to get a better idea of what autonomous cars mean for our future, I ended with far more questions than answers. How did our current traffic laws shape this technology as it exists today, and how might our laws be impacted if it becomes the dominant mode of transportation? So, to get some insight, I spoke to a top expert on autonomous driving. 
Bryant Walker Smith: My name is Bryant Walker Smith. I'm at the University of South Carolina, and I look at what I call the law of the newly possible, this dance that laws and technologies do with each other, how one affects the other. 
KM: Professor Smith told me that despite the U.S. being a complicated patchwork of state and federal regulations, driverless car companies have found few roadblocks in getting out to the street. 
BWS: Fifteen years ago, I went through all of the state laws and tried to find any examples of laws that would be inconsistent with automated driving and generally found very few. It's overly simplistic to suggest that the federal government is responsible exclusively for vehicle design, because it's not, or that states are responsible exclusively for vehicle operation or operational safety because they're not. In fact, there's lots of overlap. And so, when we're looking at all of the laws that are relevant here, it's not just laws that are specific to automated driving. It's all traffic laws, all vehicle laws at every level of government. 
KR: So, we didn't have to repeal or make any new laws for driverless cars to even get on the road. But I know some states have begun passing laws regulating the field, right? 
KM: Correct. And there are many examples of how both states and federal agencies have jumped into this conversation. 
BWS: Jurisdictions like California or Texas that have adopted specific regulatory frameworks for automated driving. The work at the federal level with the National Highway Traffic Safety Administration, or NHTSA, that has to some extent mapped some of the existing laws onto automated driving and the current efforts to develop a specific regulatory framework for them and distinguish those from the laws of general application, that the background laws that we all have to comply with in driving, and how those might apply to automated driving. 
KM: But there's not just laws on the roadway. There's also around 240 million American drivers. 
BWS: One of the constant questions has been what it means to comply with traffic law, particularly in a world where there is mass non-compliance. Sometimes in ways that are bad, speeding. Sometimes in ways that are understandable or very necessary, giving extra space to a bicyclist when you're passing them, pulling across the double yellow line to pass a parked truck. This is relevant to companies because you're asking, first of all, what should we conform to? Should we conform to law as written or kind of law as practiced in, in the driving of, of ordinary people? Secondly, it could have, particularly down the line, a lot of implications for the legal risk that companies face. One of the more difficult questions that is still being sorted out, even in a place like California, is what it means to enforce traffic law vis a vis these companies or these vehicles. And to put that in simple terms, one of the conceptual questions has been who or what's the driver of an automated vehicle? Who or what is driving or operating it as a matter of state law, or even as a matter of federal law? And we see very different answers. 
KR: How did professor Smith answer that question? 
KM: He believed the autonomous car companies should be considered the driver. Some would say the hardware and software could be seen as the driver and others take it further, claiming there is no driver per se. 
KR: OK, so simple question here. Who gets the ticket when these things make a wrong call? 
KM: It's a good question, one for which there is an answer. But under that answer lies many more questions. 
BWS: Local authorities in Texas have sent tickets to companies. California has legislation that creates kind of a quasi-ticket regime, again, for companies. 
KM: In late April, the California DMV adopted new rules, creating a process for police officers to issue tickets to driverless car companies like Waymo. These rules also require robotaxis to get out of the way of emergency first responders. This is to address past instances where first responders have claimed the cars failed to yield to an emergency vehicle. 
BWS: So we see these different ways of approaching it, and another way of looking at it is to say, well, rather than we can talk about a human person, we can talk about a legal person like a company, or we can, you know, shift to a world where that simply has no relevance. And just as there are no longer any elevator operators, it's just simply the elevator. You know, in the future, any failure would be just a defect of a product or a failure to comply with the safety regime. 
KM: And this brings us to where the federal government may assume a larger voice in regulating autonomous driving systems. 
BWS: So, this is one of the things that's actually very interesting to me as a question of federalism and as a question of law, potentially automated driving could dramatically expand the federal government's role in operational safety. If the National Highway Traffic Safety Administration, NHTSA, takes up that invitation. And let me explain why. So, today, if I get in a car with bad brakes where the lights don't work, I'm not belted in, there are, you know, kids in the back jumping around. None of that is a is a defect in the vehicle. All of that is operational deficiencies that I, as either the vehicle owner or vehicle driver, am legally responsible for on a state level. In the future, though, if an automated driving system, this hardware and software engages on a vehicle that is not roadworthy because the brakes are bad or people aren't belted in, or the lights don't work, all these reasons, then that ADS could be considered defective as a matter of federal administrative law for driving when it is unsafe to drive. And so, suddenly all of these things that previously were a question of state law become a matter of federal defect law, because the ADS should not have operated under those conditions. 
KR: As we talked about before, the presence of these cars has been steadily increasing in U.S. cities. Why do you think we're seeing such a push now? 
KM: Anyone who drives knows that safety on U.S. roads sucks. There's a reason it's called defensive driving. The National Highway Traffic Safety Administration, NHTSA, estimated that in 2025, there were 36,640 traffic fatalities. That's around one hundred people dying in an accident every day for a year. Recent years show an incremental decrease in that total number, but we're still talking about thousands of lives lost every year to something that could be avoidable. Many of these automated car companies have built their products as a means to reduce that number, but these cars still get into crashes and violate traffic laws. According to NHTSA, there have already been over 230 crashes involving driverless enabled vehicles in the first three months of this year. Last December, the Austin Independent School District reported 20 instances where Waymos passed stopped school buses. That led the company to recall 3,000 vehicles in San Francisco. A cyclist is suing Waymo after one of its cars parked in a bike lane to drop off a passenger who opened the car door into the cyclist's path, causing them to be ejected from their bike and land on another Waymo that was also parked in a bike lane. These accidents have also been fatal. In 2018, an Uber test vehicle struck and killed Elaine Herzberg in Tempe, Arizona, making her the first pedestrian to be killed by a self-driving car. Data from NHTSA has shown that self-driving cars have fewer accidents than human drivers, which has fueled these companies to push to get their systems out on the road as fast as possible. 
KR: But in this rush, could we be trading the problems of today for new ones tomorrow? 
KM: That's a key ethical dilemma that came up in my talk with Nassim Parvin, a professor at the University of Washington’s School of Information. 
Nassim Parvin: I study many things. I've written about digital fashion assistants and self-driving cars and augmented reality, but always with the lens of ethics to think about how these products that we design shape our social and political relationships. 
KM: Professor Parvin told me that traditional cars, you know, gas-powered, human-driven, are starting to run bare. And we, the public, are looking for something new. 
NP: We know that, you know, cars, as we have them now, are reaching their limits. Fuel is expensive. Our streets are crowded. When we don't want to be in cars for long stretches of time. And that ideal of freedom that sold in, say, advertising around cars hasn't materialized in the 20th century. If you go back to the beginning of twentieth century, when we had horse drawn carriages, and actually there was a problem with horses and the smell, you know, upkeep of horses. And then these cars were introduced as clean modes of transportation, but actually they made a lot of noise. They were not reliable. And we were always told, ‘Oh, but wait a minute, we are going to perfect this technology. And once it's perfected, once it has that infrastructure built around it, then it will deliver the vision of freedom and efficient access to everything.’ But the more we became reliable on cars, we began to see the side effects of that vision. We began to see how the number of fatalities rose. We didn't get rid of noise and pollution. We actually got a different kind of noise and pollution. Car companies got away with not paying for roads and infrastructure around cars. So, it's actually the public and taxpayers who pay for that infrastructure, but car companies benefit from it. 
KR: So, if this story repeats itself, driverless cars might not be the godsend these companies have made them out to be. 
KM: It's possible. And this trade-off isn't quite like the transition from horse poop to carbon emissions. Autonomous vehicles come with a whole host of concerns tied to our digital age. 
NP: I would be careful about categorizing these, I would call them entrapments, as cars because they are very complex technology, right? There's an automated system that comes from a centralized place, the company, and is controlling all of these cars. Now imagine a non-democratic government and how these cars might be used, say, to disperse protesters, or how can they be trained in order to create a mobile barricade? Now, these scenarios make it clear that if we categorize this technology, say, in the same category as, you know, weapons or autonomous weapons, a whole other set of laws begin to apply to them. And a whole lot of other set of precedent applies to them. And so, we need to be careful, even in the categorization of this technology as a car, because it can be put to other uses that cars that are owned by individuals cannot. 
KM: It's also concerning to professor Parvin that much of these cars’ software, the thing that is making the decisions, is kept under wraps. 
NP: It's striking to me that these cars just suddenly appeared on our streets. You know, the public was not consulted, and we were not told exactly how they function. So Waymo, I believe in 2022, sued in order not to have their algorithm be public. 
KM: That lawsuit she mentioned was between Waymo and the California DMV. Waymo wanted to stop the DMV from releasing trade secrets related to how its robotaxi works, obviously to keep competitors from copying them. And a California Superior Court ruled in its favor, handing down an injunction allowing the company to keep some of its secrets. 
NP: These algorithms are programmed to kill. So, in a situation where an accident is inevitable, the algorithm decides whether, say, save the lives of people in the car or the pedestrian outside. And you can imagine that such a difficult and life and death choice should be a subject of public scrutiny. 
KM: But we're talking about something that the public at large knows all too well can be dangerous. And I don't think these car companies have proven to consumers quite yet that they are 100% safe, which explains the hesitation for many. 
KR: This also reminds me of debates over social media algorithms, their methods for keeping people on the platform as long as possible, and the types of content that get promoted to certain people. He who controls the data controls the masses. 
KM: During my conversation with professor Parvin, we turned to the link between power and technology, and she shared with me this theory of the politics of our technology. 
NP: The author is a political scientist, Langdon Winner, who argues that artifacts have politics in two ways. One is that by virtue of being there, they exert power.
KM: So, an example of this would be hostile architecture, where a bench is built in a way where a homeless person wouldn't be able to sleep on it, or maybe studs on a ledge to prevent a person from skating on it. These are not necessarily harmful to people, but they show how certain behaviors can be controlled. 
NP: Now, another way that artifacts have politics, which is more insidious, is when they are so dangerous that they are not compatible with democratic modes of governance. Think about when you get on an airplane. You don't want your, you know, your airplane to be governed democratically. It's a very dangerous technology. You want somebody in the cockpit who knows what they are doing and is awake and sober. But you can imagine also how catastrophic it is if that person, say, is drunk who is in the cockpit. Now, technologies like nuclear power, or I would say self-driving cars or smart cities, are of the second kind because they tend to be so dangerous that they are not compatible with democratic modes of governance, and they require a top-down approach in their governance. Now, that means that then one person, a few people or a company has a lot of power by virtue of being in charge of these technologies. And consider this, many of the autonomous car companies that are either out on the roads or are actively testing, are already tied to some of the world's biggest tech companies. Waymo is a subsidiary of Alphabet, Google's parent company. Amazon owns Zoox, and Tesla is owned by Elon Musk, the richest person on the planet who owns the social media platform X and has a lot of friends up in Washington. 
KR: There's quite a bit of power already consolidated within that small cohort. Is this technology worth turning over more? 
KM: It's certainly a question worth considering. Now, there is another thing related to the ethics of this technology that is worth talking about, and it's probably the one that's on many people's minds, and that's labor. I'd argue that no other job has been impacted by robotaxis more than rideshare drivers themselves. But don't just take my word for it. 
Sergio Avedian: Every driver in Austin I talk to, because they do the same thing hotels and downtown and airports and things like that, they're all complaining about the same thing because I'm not getting a ping ping, meaning a request from a rideshare company because Uber has a deal in Austin with Waymo. I am Sergio Avedian. I have an interesting background. I have a finance degree from USC, and I was on Wall Street for about 28 years as a derivatives and equities trader, and then I retired, quit whatever you want to call it. So, I was researching about a book on the effects of algorithms on the human race. And I figured, what better way to figure this out by, you know, looking into rideshare in that space that's, this is about, this is 2015, 2016. Well, look where we are today, right? 
KM: A decade later, Sergio took to moonlighting as a rideshare driver to get a read on how the algorithms used by companies like Uber and Lyft impact the everyday driver working in the gig economy. In addition to working as a financial coach, he's also a regular contributor to The Driverless Digest, which covers all things in the autonomous driving space. And The Rideshare Guy, a website and podcast that provides analysis of the ride share industry. 
SA: Well, personally speaking, I am the voice of the driver in the country. I am the advocate for them. And we keep warning drivers saying, ‘Listen, you know, the light at the end of the tunnel is not the exit of the tunnel. It's the train coming at you. You need to have a plan B, plan C, this is not going to last forever. You made your money, hopefully. Hopefully you saved your money. Hopefully you invested it and maybe it's time to look into doing something else.’ 
KM: And this does not bode well for the millions of drivers who are either making a living off of these apps, or are just supplementing their income. 
SA: For the big companies, the big cost is the human, right? For Uber, the big cost is the human, for the trucking company, the big cost is the human. Human is probably 50 to 60% of the cost of doing the business. Well, they're all looking to replace the human and autonomous vehicle. Like Waymo doesn't have to take breaks. They work 22, 23 hours. They're going to do triple the trips you do in a day because legally and on the apps, you cannot be online for more than 12 hours. Legally, you have to take a rest for six hours. Same thing with truckers and same thing with rideshare drivers. And I'm going like, this thing is a machine. It's not going to make mistakes. It's not going to get into accidents. It's not going to get sleepy. It's not going to call in sick. 
KM: Sergio cautions against Big Tech's claims that these are the necessary steps to reach techno utopia. 
SA: What I'm not buying by the tech bros is that, specifically Sam Altman, is that it's going to be utopia. It's going to be universal high income. I'm like, how's that going to work? I'm not sure, you know, their point of retraining and all that stuff. I'm not buying any of that. Okay. Like if you're a 40-, 50-year-old rideshare driver, it's tough enough as it is with the macro economy to get a different job out there. I'm not buying any of that snake oil. So, we retrain and we reposition? And I'm like, ‘No, I don't think it's going to be like that.’ But then I'm also giving them the benefit of the doubt. And I'm saying time will tell, but I'm not going to sit idle and put my head in the sand and say, ‘Well, everything will be fine.’ 
KR: So, these jobs are just gonna vanish?
KM: Not quite. The experts I spoke to told me that often when a new technology eliminates one job, several others are created to fill the space. Professor Parvin gave me an example from history that might be instructive here. 
NP: Ford Model T came with an innovation in the assembly line, which was an assembly line that moves as opposed to one that stays stable. So, if the assembly line is static, it means that multiple workers are working on one car. When the assembly line is dynamic, it means that one worker is, say, tightening one screw and another worker is doing something else. Envision Charlie Chaplin's “Modern Times.” So, in a way, industrialization coincides with a different model of work that is productive for the machine, for capitalism, but is not fulfilling for the worker. I'm not so worried that we will lose jobs in this sort of, in this drive towards self-driving cars or automation or AI or whatever. But actually the kinds of work that we create are going to be kinds of work that are even less fulfilling. Say, in the social media example, think about workers across the world who have to watch horrific videos in order to create a social media platform that, quote unquote, is safe for us. So, it is still work associated with these technologies, but it's invisible work and its non-convivial work. I would want us to think again, more carefully and critically about how we think about labor and kind of how we make the labor around these products more visible in order to be able to make a more informed decision about them. 
KM: That is going to be an increasingly more important debate in the years to come. While we've been talking more directly about autonomous driving, the technology driving these cars is the same technology that's shaking up the entire world. And it's not just drivers who are headed for the cliff. 
SA: This thing called AI is replacing top level software engineers. Look what's going on as far as the major layoffs around you, right? Oracle just laid off 30,000 people. Meta just laid off 8,000 people. These are not, you know, ditch diggers. These are top level high earners, right? So, if this is replacing those types of people, I mean, what about a rideshare driver? Who cares, right? Well, that's a human. And their only livelihood is this. And how difficult is it going to be to retrain all those truckers? Hundreds of thousands, millions of truckers, right? I don't think people are just paying any attention to any of that. 
KR: For someone who does rideshare himself, what is Sergio's plan and does he have any advice? 
KM: He plans on taking the ‘If I can't beat them, join them’ approach. 
SA: Well, I'm going to call what I'm doing is MIT, not Massachusetts Institute of Technology, but Mobility Investment Trust. I'm going to buy a bunch of cars, a thousand of them, whatever as big as the fund is, and hook them up to a network and start creating income and send dividend checks out. So, I'm looking at it as an opportunity for me to increase my passive income. I just launched my website educating and teaching drivers how to invest. Teaching them about power of compounding. I launched that about four months ago. And because I truly believe your financial freedom is all up to you, it's not up to anybody else unless you have rich parents. So, you know, financial literacy is key. But my solution is, okay, I'm just joining the party at some point. If you're an electrician, plumber, AI is not going to replace you that easily. So, do something that AI cannot touch you because it's touching every part of society. It's going to touch every part of society. If you're even if you're a lawyer or whatever you are, I mean, it doesn't matter. So, all I'm saying is again, not being negative, just be ready for the change. Adapt, adjust and find a way to benefit from it. 
KM: Whether you like the idea of self-driving cars or not, they are on our roads as we speak and they will only grow in numbers from here on out. But the moment we find ourselves in now is how we, members of the public, consumers will choose to accept this new technology. We've covered a lot of the issues with driverless cars, but we also know that the status quo won't work for the next 100 years or so. So, right now is a better time than ever to educate yourself about how this technology could change your city and examine your relationship with transportation, and what makes sense for you in the future. Just don't forget, the robots can't control us if we don't let them. Thank you so much to professors Bryan Walker Smith and Nassim Parvin, and a big thanks as well to Sergio Avedian and the folks over at The Driverless Digest and The Rideshare Guy. 
KR: When it comes to driverless cars, most of us want the government to govern, to step up and make the hard decisions. But what happens if the government quits? The U.S. is made up of a patchwork of regulatory regimes, and sometimes in some places, small town governments collapse. Next time on Sidebar Amanda Pampuro brings us to Hartman, Colorado, a rural town left with no functioning government, no way to elect a new one and no one to repair the water tower. Look out for all our coverage on courthousenews.com, and don't forget to follow us on social media and all your favorite podcast platforms. See you next time.
(Outro music)