a16z Podcast - a16z Podcast: Cars and Cities, the Autonomy Edition

Episode Date: February 8, 2017

Thanks to freeways, cities became something to get through instead of something to get to. Now, as the next transportation revolution -- from rivers to trains to cars to autonomous cars -- promises to change the face of our cities, what happens to car culture, infrastructure, and more? Who owns what, who pays? And what about the design -- and product management -- challenges, whether it's designing for user trust, city adoption, or an ever-moving target thanks to constantly evolving tech? This episode of the podcast (in conversation with Sonal) covers all this and more, featuring: a16z's Frank Chen, who recently shared 16 questions about autonomous cars; Taggart Matthiesen, director of product at Lyft who covers the core platform as well as development/strategy for autonomous vehicles; and Carl Pope, former executive director and chairman of the Sierra Club -- and author (with former NYC mayor Mike Bloomberg) of the upcoming book Climate of Hope: How Cities Businesses and Citizens Can Save the Planet. Will curb space be the new shelf space? When we value the "iPhone-ness" over the "carness" of cars, what changes? And... will we all drive less, walk more?

Hi everyone, welcome to the A6 and Z podcast. I'm Sonal. Today's episode is all about cars, car culture, autonomy, and cities. Transportation revolutions have shaped our cities throughout history. We've literally built cities first around rivers, then around train lines, and then designed entire cities and suburbs around cars. But as autonomous cars come on the horizon, and as transportation as a service already becomes more widespread and ubiquitous, how do we think about what changes are coming next? What are the implications for design, product management, infrastructure, and more? Joining us to have this conversation, we have A6 and Z partner Frank Chen, who's talked a lot about AI and autonomy, including most recently sharing 16 questions for autonomous cars. We have

Taggart Matheison, director of product at Lyft, who covers a core platform passenger experience and product development strategy for autonomous vehicles. And finally, we have Carl Pope, who kicks off the conversation. He's a former executive director and chairman of the Sierra Club, and the author with businessman and former mayor of New York City, Mike Bloomberg, of the upcoming book, Climate of Hope, how cities, businesses, and citizens can save the planet. Well, I think the core question we're facing with autonomous vehicles is whether they're going to impact cities the way cars did or the way freeways did.

Why do you make that distinction between cars and freeways? Cars were essentially just carriages with a different form of locomotion. Horseless carriages. That's right. Cities accommodated them. The great era of urban mass transit came after cars, the streetcar suburbs that spread out from Boston and New York and Philadelphia and San Francisco, and Los Angeles were all powered while they were cars. They were quite compatible with cars. Cities managed cars. When Robert Moses brought the freeway in, freeways wrecked cities because cities weren't in control of freeways.

Freeways were driven through cities by state and federal agencies. and cities for the first time in human history became things to get through rather than places to get to. And autonomous vehicles can either develop as a way of making cities not a destination but a thing to be transited

or if cities take charge of them, which I'm in favor of, then we can make cities once again a place to go to and we can have an era of urban flourishing that will actually be unprecedented in human history. I find this very counterintuitive because you hear a lot of things, especially in these discussions around gentrification and cities now sprawling out into

suburban centers and going beyond that everyone's becoming more and more dispersed. What's the relationship in this email you just painted of the suburb to the city? From my perspective, we have to remember that in every American city, with the possible exception of New York, more space is devoted to sleeping cars than sleeping people. And what happens if autonomy arrives in the form I anticipate and in the form that I certainly favor, which is autonomous vehicles which are shared and used multiply,

then all of a sudden a huge amount of urban real estate, downtown central city real estate opens up. And I think what he does is it sucks both residential and business activity away from the inner ring of suburbs. Those hollow out, and then we have the challenge of what do we do now with a hollow core around a vibrant, a hollow ring around a vibrant metropolis. So I think the real challenge is going to be what happens to that first layer of low-density suburbs.

I hadn't thought all the way through to the suburb yet, but I think about what happens when you imagine all the things that will disappear if we have shared autonomous cars. your garage space can be reclaimed. You drive out onto the curb. All the cars parked in the curb don't need to park there anymore. Then you drive to your workplace, which has a garage or a parking lot, that garage or parking lot will disappear. And that place, that space can be reclaimed. And then you go home from work and then you go to a store, like you go to a grocery store, and its parking lot can get reclaimed. And so I think of all of these spaces that we can reclaim, even in cities and certainly in suburbs. And it's pretty exciting what we can do. It seems like commutes ought to get shorter

with all of the space, right? Work and life can be co-located. And there's other things that will disappear too. So the auto repair stores, the autonomous car can go wherever to be repaired. They don't need to do it in the city. The glass repair stores, right? So think of all the infrastructure that supports these cars because we're going to have centralized fleets to get the cars fixed and maintain, not my problem. So all of this space is going to free up. And if cities take control, we can be smart about what to do with that space. And I would say even beyond space, the thing I get excited about is time. Forget traffic for a second. And imagine trying to find parking in the city, the amount of time that you spend to find that parking spot before dinner, to go then find your car

after dinner. It's just a significant amount of time that people are in these cars, you know, navigating through these pretty dense areas. And how far will I have to walk? That will just disappear. There'll be generations of people who will never know that anxiety. How fast do you guys think that will happen? What's the rough time? Well, if you look at the public estimates, by the auto manufacturers. I'm talking about the incumbents now. You'll see a range anywhere from 2019 to 2024 being the year where they first introduced their autonomous vehicles. It might happen even earlier for fleet sales. The most aggressive auto manufacturers are saying 2019 is when this revolution starts. And then we have a question of what's the adoption curve? Will it be S-shaped

like it was for the iPhones? Will it be straight up to the right? Because it's just so awesome. and obviously that has implications of which companies are going to win, but it's going to start soon. Just to do a quick terminology break, could you break down the levels that we often talk about when it comes to the future of autonomous cars? Yeah, so in a nutshell, the Society of Automative Engineer says there's six levels, Level Zero through Five, and Level Zero is basically a car you completely drive yourself, steering wheel brakes. And then Level 5 car is basically a car that drives itself in any situation. The interesting level probably is level four. which is a car that will drive itself,

but only in certain situations like where it has a deep map of that environment, right? So the most people will think of self-driving cars as sort of level four, you know, I can drive in most major cities. Where does like a Tesla fit?

So a Tesla today is a level three, which is it drives, it keeps the lane, it does lane changes, it's got adaptive cruise control, it'll do emergency braking, right? So there's a lot of driver assists so that you don't have to pay complete attention,

especially on the freeway. So that's sort of where we are now. I think one of the things that's important is if we go to shared vehicles, and we probably get to shared vehicles not by autonomy driving it, but by... A fleet of people, not just a fleet of cars. Correct. But every shared vehicle replaces roughly eight other vehicles, even today with today's technology and today's ownership patterns. If you only have an eighth as many cars on the road, but people are still driving as much, more or less, that means every car. car drives eight times as far a year. Right now, we turn over the automobile fleet every 13 years,

and most of the projections people are making about how fast this happens assumes that 13-year turnover rate. I've never thought about the obsolescence question. If you only have an eighth as many cars and they're driven eight times as far and they're retired, let's say, five times as quickly, the whole automotive fleet is turned over in three years. So it might start a little later than the optimists think, might start in 2022, but it might be over by 2025. Why do you say it might be over? Because in three years, the whole fleet will have rolled over.

And once you start selling autonomous cars in large numbers, people will not buy cars that have to be driven. Being able to drive your own car is going to become a very expensive option. It's going to be like leather seats, but much more expensive. Or like the vintage Rolex when it comes to wearing a watch on your wrist. Exactly. What do you think is driving this surge of millennials? I hate using the capital M word who, I think Pugh most recently reported, have way less desire to own cars who don't even care about getting their license. When we're my age, you were tromping at the bit to get your license. They could not give a shit. What do you guys think is driving that? I don't think they're thinking, mom, dad, I'm not getting my license because guess what? When by the time I'm 25, there will be autonomous vehicles. Well, I have kids in this cohort. My oldest just got his driver's license and he had to see it to be dragged reluctantly into it. you read your 16 questions about autonomous cars deck.

Exactly. Exactly right. They're like, I'll just wait for the autonomous cars. The interesting thing is their social life unfolds much more on screens these days than in real life. So in my generation, you were desperate to get into a car because then you could go to your friend's house and play Dungeons and Dragons or go to the mall or whatever. I just had to laugh about the mall for a second because I'm remembering the scene from Mean Girls, the movie, where Lindsay Lohan, who'd grown up in Africa, sees the scene of all these teenagers gathering in a mall and they're all waiting around a fountain. And she has a vision of them going to, like, the watering hole. Right. Whereas today, all of that is mediated by our screens.

And so, like, you build social bonds via apps. And to some extent, muscle memory, I look at my phone and there's just certain apps that I trust. And there's certain mechanisms, behaviors that I do. And a lot of the, I almost said kids, because I think they are kids because I... That's fair. They're younger than us. Turn 40. And everyone else is like, hey, we're millennials.

We are all older than millennials in this room for the record. But you ask about car ownership, cars are a paint. And they've figured out how to experience life using transportation as a service versus ownership. I also think the pacing of this next generation is very different than the pacing of my generation. I can tell what the demographic a movie was made for by watching how it's edited. Really? And if it's edited very tightly, I know it was aiming at a younger demographic. Driving is slow.

Doing one thing for 15 minutes. It's like agonizing if you're used to being able to multitask on your screen. And so I think that driving is just a drag. It's like if you take in kids my age and said, oh, by the way, yeah, when you're 16, you can get an ox cart. They would not have been so thrilled to get an ox cart. And for this generation, the automobile is an ox cart if you have to drive it. Is it really an ox cart, though? Because at the end of the day, an ox cart was a utility.

and a car and automobile was a status symbol. There's a whole glorious car culture and a lot of back and forth about whether that car culture still exists and part of this might not be distributed evenly. This could be concentrated in certain regions and diffused in others. What are your thoughts on that diminishing or persistence?

One of the things I watch to sort of figure out if there really is a culture of cars is watch consumer reports reviews of cars over the years. Ooh, good trick. And basically, in the old consumer reports, It was all about the goodness of the car, horsepower of the engine, skid pad performance, breaking distance. Now, if you look at consumer reports, like, they spend a lot of time on the infotainment system. What is the user interface of the car?

The negative reviews will be for, hey, this is a great car, but the infotainment system is terrible. It takes two seconds to boot, and it's three seconds to, like, register presses, and don't buy this car, it's a terrible infotainment system. And so, like, if you just watch that trend, because that wasn't true before, we didn't rate the stereo UI, right? It didn't really enter into the conversation. I think these days, it absolutely enters into the conversation. And so the carness of the car is less important. The carness of the car is very important, right? What's the interface?

And I think we have to remember that for my generation, the car was the first status symbol you could own. For kids today, when they're six, they start competing for who has the best electronic objects. So they actually have focused status on handheld electronic objects, whether they their game consoles or phones. So the car is like, eh, do I really want that status? Right. Yeah. It's much more important whether you own an iPhone or an Android phone and what set of apps you use

than whether you drive a Ford or a BMW or whatever. So given that shift, what does that mean for the current and future of designing these new cars? Essentially, these cars that are no more software than they are people caring about the mechanics. Well, one of the big changes, especially if we're going to, to switch to fleet ownership is the car manufacturers will start being a lot more like Boeing and Airbus than they will be like Ford and BMW today. So if you think about Ford and BMW, they spend a lot to establish a brand persona in the buyer's mind, right? BMWs, the ultimate driving machine you hear on the radio, you've got billboards. Boeing and Airbus don't do that because they know that who buys the

plane is the operator, the fleet operator. It's United and Lufthansa. And so they market to them. Right. So the things that matter to people in terms of differentiating in a car will no longer matter. Basically, they need to win lift an Uber. So in a very simplified way, it's like a shift from like a B to C to a B and just really thinking about the enterprise consumer. Yeah, that's exactly right. So if you think about the sound the trunk makes when you close it, like that's a B to C thing, right? Like, oh, I love Mercedes because the sound the doors make when they close. Solid as a bank fault.

But the safety and the wear and tear and the brakes is probably way more important to you. like the person you might be thinking about buying, if not building a fleet of said cars. That's exactly right. Will the cargo 300, 500,000 miles is going to be much more important buying criteria than what sound does the trunk make when it closes? I can accept that millennials have given up their desire to like tie their identity to their cars. But given that it's so ingrained in our culture, how do we now think about designing for

a culture where that car is no longer that container of identity? I would almost argue that we can keep that identity. identity. So yes, I mean, when I was 16, I had a, or 18, a 1970 suburban. My identity was this massive thing and it was awesome. And when I parked it at high school, everyone looked around like, wow, that thing is a beast. I think you can do the same thing where when I get in that car, I can still have my identity. And that's where I think the personalization of understanding who that passenger is and what you're trying to get out of that ride. It could be that a couple friends of mine were going to a party and it's going to be in Taggart mode. And Taggart mode is a certain

instead of songs, or maybe it's lighting, how do you keep that passenger connected? How do you keep providing a personalized experience? What would an analogy be to an existing consumer thing that we have today? I think like if you look at Spotify. I was looking at Netflix myself. Exactly. Exactly. So like Spotify or Netflix, it's always there. You've got your personal history in terms of your playlists or the movies that you've seen. Imagine if you take so many rides, we know a lot about you. And I think the goal there is to make this a really person. personalized experience and working with car manufacturers to reimagine what that looks like. So rather than just getting in the back of a car, how do we take someone who's in a meeting

right now and extend that meeting throughout that trip? Like the car just arrived. You got in. It was your car. You continued with that conversation and you got out and you were able to complete that without any hesitation. And it sort of seamlessly doesn't interrupt. What you're describing is ubiquitous computing plus context-aware computing, which is another vision that scientists have been talking about for years, where you do have this ability to have contacts persist with you. Like, why are our phones dumb? Our phones should know so much more about where you are in a meeting right now. Why wouldn't our cars know? Why wouldn't other things in our environment know? Ford is already making cars that have more code in them than a 757. But I think something else

very interesting is likely to happen, which is I think cars become very modular. I think you have a basic chassis and drive train, and then it can be easily outfitted and perhaps outfitted by the fleet owner to be a really small, it takes two people to work, to it's for a date night, and the fleet owners will actually start marketing the fact that when you call them up, you won't just say to lift or Uber, give me a black car or something simple, you'll actually and are 14 options that you want in the vehicle that comes to you. I'm thinking of like the celebrity requests when they're like, I need only green M&S and I want my vitamin water.

Everybody will be a celebrity. That's a great way of thinking about it. And that's what will be marketed. That will be the value proposition is we customize your date car or you're going to work and you want four people and you want a certain set. You want Apple not, you know, some other competing software. GM's got a really interesting product experiment in this direction. So they announced something called Book by Cadillac. And basically, it's a $1,500 a month service that allows you to rent any Cadillac you want. So you could get the stretch limo or you could get the, you know, whatever. They have sports utility vehicles. They have two-seaters. And so it's basically $1,500 all in. You rent the Cadillac you want. I don't care to own a specific vehicle because that specific vehicle isn't an extension of my personal identity anymore, right? Or past that. Right. But Cadillac's true. trying to say, but we want Cadillac to be the brand that you're associated with. So let's not

talk about model identification. Let's talk about brand identification. So this is interesting because you're bringing up the as-a-serviceification of this model. It's really more around, there's an app, there's an experience, like Spotify and Netflix. You're not actually owning the vehicle. You're not owning the CDs from a Spotify. You're not owning the DVDs on Netflix. Correct. And that's kind of where we are today with transportation as a service. It becomes a very powerful way to get around versus the traditional car ownership. You know, I think of subscription models, which I agree that so much of our world

is moving to as a service for everything. Are people going to end up paying more money for cars over time as a service than they will for if they had bought a car? Well, I think it's an important question. If GM is doing $1,500 a month all in, right, compare that to any time I want to go somewhere and you'd have to do the math

and sort of figure out when your bric even point is, we're starting to hear about this exact same trend for bikes, right? Which is we have the city share bikes and then there are more aggressive startups that are doing bike sharing services and you have to sort of do the breaking point. How much would it cost for me to own and operate that bike? And that includes maintenance and everything, all fully loaded. It can't just be the object of buying it at point of purchase. Yeah. Today, it comes with a driver, right? So the experience is fantastic. It's having an on-demand chauffeur anywhere you want to go.

At some point, when the autonomy comes into the equation, then it might get close to the price at which you sort of own and operate and drive your own car. and the price of your apartment will go down when the apartment building no longer has to provide a parking space. Or what other people do in cities like New York and San Francisco, which is you have to actually rent a parking space. That in other cities, the price of that parking space is literally the price of the rent. So, three to $500 a month for your parking space,

and that drops out of your budget. And there is the time that you spend looking for that. And let's be clear, what's really valuable to people is the ability to use the time they currently spend watching the bumper ahead of them, which is not a high value experience. Watching the bumper ahead of you is not a high value experience. This is why I believe our podcast is really popular. I mean, there's been a lot written about how podcasts have had a resurgence.

The technology is not new. There's probably more now than there was before, largely because there's so many smart connected cars. And people have, quote, idle time in the form of traffic. That's not counting idle time of cars just sitting idle, of course, which is its own statistic. Great as your podcast is. There are a lot of drivers who would prefer to be sitting on their laptops and doing work out. Yeah. Or watching a movie. Okay, so one of the thing that's interesting is you have these original equipment manufacturers who just hand off like this chassis and people can design the hardware and they can customize, but the software customization happens at a different level. How do you think about when you have this world of transportation as a service and you're literally handing off what?

The problem space is massive, right? Like that car needs to work in a sandstorm in Egypt and has to work in a snowstorm in Juneau, Alaska. It's a massive problem space. And the challenge of an autonomous vehicle being able to navigate all of those potential areas is very, very difficult. I think the interesting aspect here is when you combine autonomous with transportation as a service, what we can do is we can slowly roll that out. We know where the start and the end of the ride is before it happens. And so what that allows us to do is really constrain that problem space. And so if it is a ride or a request in an area that we're comfortable in terms of safety,

in terms of validation with the autonomous vehicle, then we'll go ahead and dispatch that vehicle. But if it's raining or a certain time of day or something else that puts the probability at some lower level that we're not comfortable with, then we just dispatch a typical traditional driver. What does that mean for geographic localization? We've talked on this podcast about what the new localization is.

And I always think of when I think of autonomous cars is I don't think of these beautiful U.S. American freeways that everyone follows rules and stands in lines. I think of India where you have like bulls and cows cross the road and food, vegetable sellers and people driving like crazy. It's insane. I don't know else to describe it. Yeah. The Baidu chief scientist of their autonomous unit says, you know, if you think you have a challenge in America, come to China and watch the pedestrians and, you know, the actual traffic. And so the localization challenge for the car manufacturers, are they going to have to ship different versions of the self-driving cars for Budapest and Boston and Shanghai? or will the cars learn the local variance in, say, the first month they're on the road? And so both research approaches are active.

I've talked to people pursuing the, look, it's going to be localized in the safe breaking distance and how aggressively to take that right turn when you're at a stoplight. All of those things will be programmed for your city. And I've also seen manufacturers saying, nope, we're going to have a general set of learning algorithms and they'll learn the conventions on the fly. Can you talk about some of the product management challenges around what you do? It's one of those products where you don't actually know what the right answer is. The technology isn't fully baked, and you're kind of designing for the future that isn't here yet.

And so there's this huge leap of faith that we are in right now, but the beauty is that we actually have a lot of the data. So we have tons of customer data. We've done plenty of user research. It was not an autonomous vehicle, but we stated that it was an autonomous vehicle for some user research. And one of the interesting things was we had a button that basically said start ride. And we said, oh, you know, the car's autonomous. You don't have to worry about anything. And almost every person would turn their head to look to see if any traffic was coming

before they started that ride. Fascinating. Trust is one of the key things that we have to capture as part of this experience. And so I think while it's an extremely challenging problem. And you got moving target, moving technology, are evolving. Everything's moving. there's a ton of tools that we have today that allow us to at least vet a lot of the experiences that we think people will have. And I think the other piece of this is that we, you know,

we are humbled. We're not looking to just flip this on in every single city on every single street. We want to do this in a very cautious and careful manner that allows us to ensure that it's the best possible experience and an extremely safe experience as well. I mean, I'm glad you bring that up because there's a tendency sometimes when I think of software eating the world for, and It's not a bad thing because it's what makes entrepreneurs to sort of just attack the problem. And that's great when you have software, which is this abstract thing. But when you're talking about physical objects and trust and livelihood, those are all pretty critical when it comes to designing a literally a moving object. And it's not a drone that it's unmanned.

Right. And safety is the first thing that we focus on. How about environmentally? Like, do we have to rethink some of these, a lot of cars have been designed for laws and regulations around environmental emissions? I have to do that annoying. In my old car, he said, do that annoying before I got a Prius, like smog test. The key intervention here, which I'm hoping cities will guarantee, is to make sure that autonomy comes with sharing, with fleet use. Because once that happens, that drives you for a whole host of reasons to electric vehicles.

Once you drive to electric vehicles, the vehicle doesn't emit. You may have to make sure the power system doesn't emit, but the vehicle itself is fine. And driving it to shared also solves the problem, which some people have worried about, that, oh, if it's so comfortable to drive, people will drive further and more hours a day. And if there's an opportunity cost, because you don't own your own car, but you're paying twice as much to drive twice as far, people will take that into account. So we view this as a crucial environmental opportunity, but it is an environmental opportunity, which is triggered by fleet and by sharing. So that's a critical component. And that's a critical component, which again, cities, I mean, there's already an interesting conflict going on. I mean, some cities have said, if you want to have curb space at our airport, you're going to have to drive an electric vehicle.

So cities have power because they control what's now become the valuable commodity. The valuable commodity is access to the customer at the curb. Curbspace is now valuable. Curbspace is essentially given away almost for free, and that'll stop. that'll change advertising. That'll change everything. So what does it mean for infrastructure when we're thinking about how when I picture that bumper on a freeway? I picture a classic freeway, which really is just a path that has nothing on it. What happens to infrastructure as a result of this revolution evolution. I think as these cars evolve, the sensors that are on these cars and the

amount of data that we're picking up allows us to basically get this heartbeat in the city of where all of these cars are, what the traffic patterns look like. And on a long enough timeline, you can actually leverage that data to start to predict. And so I think what gets really, really interesting as cities evolve, is you have a ton of this data that you can start to evaluate to determine what's the best path in terms of traffic. And so whether it's using smart signals or potentially even smart lanes in terms of lanes, certain times of day for cars that have greater than, let's say, two or three people in it, there are ways that we can leverage this data and look at these patterns to basically evolve these cities out of the congestion that we see today?

I think it's much more radical than that. Which is? When cars arrived, we had to find a way to substitute something for the horse's unwillingness to drive somebody down. Traffic lights, crossing lanes, no jaywalking laws. It didn't work very well. We still have a lot of pedestrian fatalities. We'll reach a tipping point when, say, 70% of,

of the vehicles on the road are being driven by a robot, which like a horse will not run someone down. And at that point, pedestrians are going to start crossing streets the way they did in 1900 without looking and totally focused on their screens. And at that point, we will have to take the right to drive away from people unless they're in dedicated, you can drive it lanes. We now have bike lanes which are dedicated and special. You can drive your own vehicle lanes are going to become dedicated and special in the city of the future because the risks of having driven cars with properly managed robot cars is just going to be too high. You've actually, I think, argued, Frank, that when will cars having a license be illegal?

Yeah, that's the full radical version, which is that the autonomous cars are so much better drivers that as a society, why would we let the drivers who are likely to run people over as opposed to the robots, drive at all. Now, I got a lot of criticism on that when I published the provocative idea, because it is provocative, about, wow, all-out war on human dignity and, like, why would you take away this freedom? And, like, I love driving and so on and so forth. And I think when people think about that, they're kind of thinking about, you know, sort of Tracy Chapman, fast car, you're going down the freeway, the top is down, the beautiful sunset. You are not thinking about gridlock traffic on I-5. I think La La Land is actually the first movie that actually incorporates that meme.

In La La Land, driving is not a symbol of freedom. Driving is slavery. In fact, the movie opens with them singing to escape the confinement of the traffic. That's correct. They all just hop out in the middle of the freeway. What about other aspects of infrastructure? What about parking meters and parking garages and parking spaces go away? And we can think of a million things.

What would fill that? Well, hopefully in the urban dense areas where housing prices are so high, Housing will fill those areas. Oh, yeah, right. Without the not in my backyard people. Yeah, that's exactly right. Now, that's not going to happen everywhere because, you know, housing is not scarce everywhere, in which case we could think about urban farms or we could think about workplaces or we can think about more sort of local services where, you know, you really want somebody close to you. So hopefully the urban planners will be creative on how we use this space because there's going to be a lot of space coming available. The healthiest cities are measured by

their walkability. And how would this transportation infrastructure and ecosystem change walking? What people even want to walk when you can just literally hop in like a continuous network of essentially a public transportation train, but that's really just a bunch of cars in a straight line? Well, based on where people walk today, we don't have to worry about that because people in the city like Manhattan walk a lot, even though there are a lot of alternatives. People like to walk if they don't have to compete with cars and if the experience is pleasant and one of the things we could devote that extra space you talked about to is making walking pleasant. You look at how many people in New York take a subway from uptown down to Chelsea to walk on the high line. I totally agree

with that point. I think once you kind of reimagine what it can look like without all of these cars, you're definitely going to find people walking more. I mean, it's a perfect example is the waterfront in San Francisco once they took down the freeway. And you look at the ferry building. And I take that walk and you can see my eyes get big here. I take that walk every day from the ferry building down to 185 Barry to the ballpark. Why? It's beautiful. And I think like if you start to see some of these changes in these cities, people are going to want to experience that city. They're going to want to be connected. They're going to want to talk. They're going to want to go to the shops. Right. I don't have to worry about crossing three, four layers of traffic to get to the other

shop designing around the people, not the vehicles, and that's a really important shift. I love what you said about. You could actually explore and experience the shops. Right now, shops are oriented around these walkways and like shopping malls or little pockets or neighborhoods. The mall obviously emerged as a result of the sprawl and car culture moving away from train culture. What do you guys think about what the next mall would look like? More windows. I think it's a very challenging question because simultaneously with the autonomous revolution, we're also seeing the online shopping revolution. So the question is, what do we still physically buy? So I'll share one of mine, which is I was at a nail salon in my street in San Francisco, getting my nails done.

And I realized that any time I go, there's always people there. And when they're not people there, they'll be like, wait here. And within two minutes, five minutes at most, people show up. And I'm just like, how do they do this? I've just been fascinated by it. And then I've found out that the woman who owns a salon has a couple of shops on the street. And what she does is she transports people between the shops based on their availability. And what's fascinating to me about this is that it inverts a model that I've always thought about when it comes to cars. It's not just decentralizing the driving and these different, this network of nodes of cars, but it's actually now the reverse where it's actually bringing people to all these different

places. And now you're kind of reallocating labor in a way that sort of moves with where things are. Well, that's an interesting concept, because if you think about the medical house call, if a doctor could take a car, while he's in the car going to see a patient, he's doing his paperwork and his other things. The doctor spent a lot of their time doing paperwork. Then he visits his patient at home. Decentralized. Then he goes to his next patient and does the paperwork from the first visit. You don't need a doctor's office. Exactly. That's what I think is super fascinating about this. I think about the future of commerce here. And you could equip some cars with the necessary basic infrastructure that a house-call doctor needs.

Even before we move to fleets and autonomous, the doctor in a car service already exists, eventually we'll have the robots drive. Well, so the other aspect of this, besides some of the sociocultural factors, like people's mindsets and attitudes, is the socioeconomic. And what automation itself means for jobs.

It's part of a larger horrendous problem, which is automation is replacing most analog skills. Most of the things that we do, they're somewhat routine with our hands can now be done by a machine, and increasingly even things we do with our brains can be done by a machine. You asked about infrastructure earlier, and I think one of the big opportunities to counteract the fact that we're going to lose the largest profession in the United States, which is driver. That is the largest profession. Professional truck drivers or all drivers? All drivers. That's going to go away. Where we have the

opportunity to replace that analog work is actually in taking the infrastructure. and the cities and making them incredibly wonderful. Building the high line, turning it into a park, and that requires social investment. So there's a big political question here, which is, can we grow the public budget? Because that's

where the jobs are going to be. I mean, on the one hand, yes, self-driving cars will take a lot of driver's jobs. But after the introduction of the ATM for banking, the total number of bank tellers in the country increased. Right? Because what we did was we built banks with ATMs, inconvenient

places and put tellers in them. So we expanded the reach of banks so that you're, you know, two blocks away from any banking center, whether that's ATM or one with a teller. So it's counterintuitive. You would have thought ATM wiped out all of the bank tellers, but actually the total number of bank tellers increased because we went on this expansionary, you know, sort of put a bank two blocks from everybody. Yeah. And so I think we have some of those trends yet to play out. Think of all the jobs that didn't exist 10 years ago. Like social media consultant, podcaster, right? All of these things that did not exist 10 years ago. got created because the human capacity for novel products and services is essentially infinite.

People will invent new things that they want, and technology will give us efficient ways to create those new products and services that people want, and it's going to unfold in unpredictable ways. There is an opportunity for some services to actually evolve from this. So you think about fleet management. As we evolve into fleet management where you have these massive fleets, we're going to need people to take care of these vehicles. And so I think very similar to other, if you look at the history of technology, as people

People get displaced. Other services start to pop up as they map back. So I think there's an interesting opportunity. Yeah. And I think it's very large because let's be totally clear. Right now, most cars are driven kind of dirty. Once you're driving fleet cars, they're all going to be absolutely clean. They're going to be cleaned after every trip.

The funniest thing I've ever seen around the biggest argument against the sharing economy with cars, someone wrote this tweet that the car they get into always smells. And that is a single biggest roadblock for not. not having a good experience or having a good experience in the car you're sitting in when you're taking a car. You're going to have to have cleaners. You're going to have to have mechanics because you're not going to want to take them to the dealerships. You also need to figure out what you do with the cars when they don't have people in them. So like pre-positioning to do they idle? Do they just loop around any mile is expensive in terms of an autonomous vehicle? So you need to optimize every second.

Right. That's great. Okay, everyone. Thank you for joining the A6 and Z podcast.