a16z Podcast - a16z Podcast: When Software Eats Cars

Episode Date: August 25, 2015

After the smartphone, what business has the global scale (in terms of people and profits) that make it attractive for tech companies to turn their attention and capital towards? The answer, argues a16z’s Benedict Evans, is the automotive industry. Sure, cars are mobile -- but what do our smartphones have to do with our rides? More than we might expect. Evans offers his vision of the future of cars, and perhaps the future of today’s biggest technology companies. Who will build these cars, who will own them, and who (or what) will drive them? We discuss this and more in this segment of the a16z Podcast. The views expressed here are those of the individual AH Capital Management, L.L.C. (“a16z”) personnel quoted and are not the views of a16z or its affiliates. Certain information contained in here has been obtained from third-party sources, including from portfolio companies of funds managed by a16z. While taken from sources believed to be reliable, a16z has not independently verified such information and makes no representations about the enduring accuracy of the information or its appropriateness for a given situation. This content is provided for informational purposes only, and should not be relied upon as legal, business, investment, or tax advice. You should consult your own advisers as to those matters. References to any securities or digital assets are for illustrative purposes only, and do not constitute an investment recommendation or offer to provide investment advisory services. Furthermore, this content is not directed at nor intended for use by any investors or prospective investors, and may not under any circumstances be relied upon when making a decision to invest in any fund managed by a16z. (An offering to invest in an a16z fund will be made only by the private placement memorandum, subscription agreement, and other relevant documentation of any such fund and should be read in their entirety.) Any investments or portfolio companies mentioned, referred to, or described are not representative of all investments in vehicles managed by a16z, and there can be no assurance that the investments will be profitable or that other investments made in the future will have similar characteristics or results. A list of investments made by funds managed by Andreessen Horowitz (excluding investments and certain publicly traded cryptocurrencies/ digital assets for which the issuer has not provided permission for a16z to disclose publicly) is available at https://a16z.com/investments/. Charts and graphs provided within are for informational purposes solely and should not be relied upon when making any investment decision. Past performance is not indicative of future results. The content speaks only as of the date indicated. Any projections, estimates, forecasts, targets, prospects, and/or opinions expressed in these materials are subject to change without notice and may differ or be contrary to opinions expressed by others. Please see https://a16z.com/disclosures for additional important information.

The content here is for informational purposes only, should not be taken as legal business, tax, or investment advice, or be used to evaluate any investment or security and is not directed at any investors or potential investors in any A16Z fund. For more details, please see A16Z.com slash disclosures. Welcome to the A16Z podcast. I'm Michael Copeland. After the smartphone, what business has the global scale in terms of people and profits that make it attractive for tech companies to turn their attention? and capital towards. The answer, according to A16Z's Benedict Evans, is the car business. Yes, Benedict is our mobile expert, and sure cars are mobile, but what does your iPhone have to do with your ride? More than you might expect, Evans says. Evans offers his vision of the future of cars, and perhaps the future of today's biggest

technology companies, who will build these cars, who will own them, and who, or what will do the driving in this segment of the A16Z podcast. Benedict, welcome. Hello. Let's talk about what cars on mobile have to do with each other, what cars in the smartphone industry have to do with each other, and what even brought this subject to mind for you. Well, what brought it to mind is that Apple and Google are getting into the car business, and Tesla is in the car business as a technology company, and Uber is in the car business as well, and obviously we have an investment in Lyft as well. And so cars and technology are sort of colliding in a way that mobile phones and technology collided seven, eight years ago and other things have collided in the past. And it struck

me there's kind of maybe four separate building blocks to think about that kind of linked together and sort of support each other. One is electric cars. One is the rise of on-demand services. one is the rise potentially of sort of autonomous self-driving cars which kind of come out of both of those things and sort of feed into both of those things and then one which is perhaps slightly shorter term is sort of the software in the car and how that becomes something that isn't made by the car company anymore

perhaps and becomes something that gets into the smartphone ecosystem and all of those kind of interact with each other well let's take those in turn then and start with the electric car and how that sort of lends itself to technology and a technology play a la Tesla and others? Well, the thing that is sort of most obvious and visible that people think about here is the environment and renewable energy and pollution and so on.

I think equally interesting from a technology and engineering point of view is that you change the mechanical complexity of the vehicle when you go to electric instead of internal combustion engines. and you change and so you don't have a transmission you don't have a gearbox you don't have all those moving parts anymore and so that changes potentially things like maintenance

but it also changes kind of the complexity of creating the thing and it changes what the supply chain might look like it changes what the capital structure of making cars might look like and so a little bit like what happened with PCs or smartphones 10 years and 20 years ago you move to a point that it may be easier

for people who are not in the car business to be in the car business because the amount of stuff that you have to make how that stuff gets made, how it gets how the value chain around that looks gets much much much simpler. Right. Obviously Tesla is the best example of this but the question then becomes

where does the value in an electric car lie and for Tesla it's clearly in the battery for example. Yeah, it's the battery instead of the engine clearly and but there's a sort of a longer term point which is well okay so who owns these things and how

that they move around, which is where things like on-demand and self-driving cars come in. And, of course, you can have those completely independently of electric cars. It just makes it a bit easier for some things, but you can have a self-driving petrol diesel car or steam-powered car if you want to do. And I think those are kind of separate but linked together. And the thing with on-demand, is sort of widely discussed now, is that you're going to reach an equilibrium point of the supply of drivers that gets drawn in by the price. that will drive in riders

and the number of riders that get drawn in and that will be some percentage of people in each city and it will be different in each city depending on the topology on the layout of the city and there's calculus between owning versus

between owning and owning a car versus just calling a car versus leaving a car at home and whether you commute with it or only use at the weekends on whether you have one family car instead of two all of those kind of things

it's going to change car ownership it's not going to change it completely but it will change it and it will change it kind of different degrees in different places. Now, what's interesting about self-driving cars is how much they change that. And so just as electric cars are not just about pollution,

self-driving cars, I think, are not just about accidents. And I think the way to think about that is kind of imagine your scenario. You go out with your family to dinner in a busy city, and instead of having to find a parking space, your car drops you off. and so you don't think you didn't and previously you might have taken public transport because it was so hard to park but now your car can just drop you off and there are no cars parked on either side of the road says twice as much one for traffic of course and then when your cars dropped you off where does it go does it drive outside the city centre to a cheap car park instead of the expensive car park around the corner or does it spend the next two hours driving people around for money and then pick you up later if so and so the point is what self-driving does is it transforms the layer of of the city because things like parking and gas stations and strip malls and the kind of the whole layout of the retail and commercial real estate starts to change because it's all built around parking as much as it's built around driving you have to kind of look at a kind of

a satellite map of any American city and look at all the parking space and so that all starts to reshape but then the ownership changes because suddenly you have vastly more cars available for an on-demand service. And those cars don't have a driver, and the insurance is probably costs half as much because self-driving cars don't have accidents in the same way, and the error rate will be lower. And so the price of those self-driving, of those on-demand services will be lower. And so the equilibrium point changes again. And then you have all sorts of, and so the point is that self-driving cars, therefore, make on-demand services much, much, much bigger, and they tend to result, therefore, in many fewer people owning a car.

So that's my question, then, like, who owns these self-driving cars? Is it sort of the Airbnb model where I'm at dinner and I, you know, send my car out to make money while I'm eating? Or is it this fleet kind of a view of the world where some other large organization, whether that's Google or somebody else owns these cars? Well, again, this is kind of a question of where the equilibrium point settled down. So on the one hand, you could argue, well, everybody will have a car and they'll all drive people around, so there will be no like fleet of independent cars. It will just be a person's car. I don't think one would go that far.

But the other argument would be, well, why would you need to own one if there were all of these out there? And of course, if everybody said that, then someone would have to buy a fleet of cars to service that. So there's obviously there's kind of a cycle in there somewhere. Now, the challenge in here is if there is a fleet of on-demand self-driving cars, who buys them and how do they get bought? Well, right now, well, is it going to be an individual person who takes out a lease and sends their car out to earn money? It's obviously not going to be the driver because there is no driver. So is it owned by an individual or is it owned by somebody with a big balance sheet, buys 500 of them and has the right financial structure for doing that. And once you, and that could be both on-demand and on-demand self-driving.

And once you go there, well, how is it that those cars are getting chosen? They're not getting chosen based on how nice it feels when you slam the door. and, you know, the diamond cut, chamfered edges, they're getting bought in the same way that Hertz buys cars, and they're getting bought in the same way that a big company buys PCs. That is to say, they're bought on, you know,

criteria that are not much to do with ease of use or flair. It's beigey and boring, yeah. Yeah, well, not completely beige, but, you know, how easy it is to open the case and change the network card or whatever it is you need to do these days. So, and so that becomes interesting for Apple and potentially for Tesla. because it means that, you know, if the drivers, if the people riding around and not the people

choosing these things, then maybe Apple strengths don't really get to come into play because you might not be choosing it. Now, obviously, there's kind of variation within that. So you pull out your phone and you order, you know, you order the cheap one or the expensive one or the Apple or, you know, maybe you get a choice of cards when you pull out, when you order it. But on the other hand, if you order like an Uber black car, you don't decide whether it's going to be a Mercury or whether it's going to be a Lincoln or Cadillac or the Mercedes. You just order that and what comes comes. So maybe we will see. Then there's kind of another point which is to do with the redesign of cars, of course, which is if you remove, you know, obviously

electric changes to design, but if you remove the mechanical controls, the manual controls, then you can sort of fundamentally redesign what the inside of a car might look like and how it might be structured. I mean, right now when you look at a Tesla, you wouldn't be able to tell that it doesn't got a petrol engine inside it if you didn't know. You know, it's still got a big engine compartment in the front. There's just no engine in it. And so when you go to electric and self-driving, you can kind of think quite fundamentally about how you design a car. You might also think about things like crash tests. Like if there's no, if the accident rate changes or the accidents look different, then the way you design the car might change as well. Yeah. I mean,

you can imagine it again, changing the space and the layer. I mean, it's really an interesting problem or opportunity. Yeah. I mean, people don't need to face it. forward. Right. And you don't need to have all that space in the front for the, for the, for the dashboard. Right. Tired at the end of the day, you know, you're lying down in the car. Yeah, exactly. And we won't take that conversation any further. Yes. But then you get, to the point, again, you get all kinds of sort of second order effects. So, you know, I think Carl Sagan said that it was easy to predict mass ownership of cars, but hard to put it, Walmart. Right. So what, what do you see then? You've already

described empty streets, but. Yeah. You look at photo. Look at five. photos of the world from the 1930s and all the street of your own city from the 1930s. And you'll see all those streets where you know there's a car park every five feet all the way along the road. And there's no cars parked. And the roads are twice as wide. The other thing that someone mentioned to me on Twitter was cycling is if you could cycle every day and know you would not get hit by accident. I mean, if you fall over, then it's your fault. But you would know that no one would hit you through carelessness.

Maybe more people would cycle. If you had no manually driven cars on the road, then you don't need stoplights anymore. And the cars can just stream through all the junctions, stagnant before and behind each other. The first time that that happens, and you as a passenger go through that there's a video that you have on your blog post, which everyone should watch, but it's a frightening game of Frogger where nothing gets killed. Yeah, exactly. It's like watching a circus act or something. It's like, you know, people whizzing past each other at 40 miles an hour, five feet apart. You know, crossing each other at right angles.

And, of course, if it's all moved into software, then you can do that. So you get all kinds of changes in how you'd lay out roads and how you would lay out, you know, cities and where you would think about what retail would look like. Cycling, I think, is interesting. The impact on people's time is interesting because, you know, presuming for the, you know, apart for anything else, you might live in different places in the very long term, like on a kind of a multi-decade view. But in the meantime, if you're driving for 20 minutes every day but you're not having to look at the wheel, then what are you doing? well, you're probably on your smartphone, which is cool. So the time spent online or time spent looking at the internet changes. Because nobody's doing that now.

Yeah, of course, no, nobody at all does that. But on the other hand, what happens to radio? If half of all radio listening is happening when people are driving, well, if you could be doing other things, maybe you might not listen to radio. So you get all these kind of weird kind of unpredictable effects as kind of the balls bounce around on the snooker table, on the pool table, and kind of bang into each other. And then you also, of course, there's a sort of the big question is what happens to public

transport and where does that leave people on low incomes? Because if you're kind of cherry picking people off public transport routes, then you're going to have buses that are emptier and emptier and the cost of those buses becomes harder to support. On the other hand, self-driving cars get very, very cheap. How cheap? If you are someone on a low income and you're commuting to work on public transport and it takes you an hour and a half and four buses, well, maybe self-driving car might actually be a better option for you. It might even cost the same. You know, you don't really know. So there's certainly kind of questions in all kinds of different directions. But there's also nothing. There's nothing then to prevent self-driving buses and all

or sort of, you know, a fleet of transportation being owned by government for them. Yeah, yeah, but I don't think, I mean, self-driving, I mean, yes, not having a driver in the bus means it, you know, it could participate in these kind of high-speed junctions and things. I don't think a self-driving bus is kind of transformative in the way the self-driving car is because there's always got to, you know, that's, you know, the space taken up. by the driver isn't really the point. But no, what I was more thinking is, you know, if you live somewhere that's served adequately by a bus,

but not terribly well, if you've got to walk 10 minutes to get to the bus stop or 20 minutes to get to the bus stop, then you might never take a bus again. Right, I see. And so the bus routes, as when you get out of dense urban areas might empty out to some extent. Bus routes in dense urban areas might fill up

because they might be more efficient because you don't get traffic jams anymore so you can get on the bus and it goes all the way down that busy street without stopping and then lets you off where you want to go. So you get all, I suppose, the kind of conceptual thing that we're looking at here is really shifting road transport

from being sort of circuit switched in a sense. That's not a perfect sense, but really kind of being circuit switched to packet switch switched or certainly moving public transport, integrating public transport with cars into a packet switch model. Well, let's push on that circuit versus packet example. and because that then gets to who's best suited to be doing this.

And so when you say circuits, it's sort of this manual, I'm driving, I'm, you know. Yeah, I mean, it's not perfect analogy. Maybe you could argue it's like it's TDMA in effect or everyone's kind of got their time slot. And you move to wide-bound CDMA. Like, why would you have lanes on the highway anymore? Right. Why would you not just have kind of swarms of cars streaming down the highway without stopping in any, any positioned in any place that you like on the road,

just as long as there's five feet between them or two feet between them, depending on what you want to do. You could start having pelotons of cars. So you've got 15 cars all going down the highway, two feet behind each other at 150 miles an hour. It sounds like an aspect. So who is it? Who is it?

Yeah, exactly. And so then who is it, round and round and round and round. So who is it that builds? Well, this is the, again, I mentioned kind of one sense in which is it's potentially challenging for Apple in that it's, you know, who is it that's buying these things. But there's another point which is, which is the, what's the hard part of the software here? there is a hard part in the not hitting things in actually the car itself but then there is all the kind of the routing and the load balancing and the optimization and the prepositioning of all the cars because it's one thing to get into your ubu or a lift or you'll get taxi or whatever and it to summon one and the nearest one comes to you it's another one at kind of five in the morning all of those 150 or 500 or a thousand automated self-driving cars leave their charging stations where should they be at 7 o'clock in the morning when people start summoning them where should they be at 730

where should they be at 11.06 and where do you move them and which car do you send to take which person to which route and what do they do on the way and where do they go after that and so that kind of mass positioning of hundreds or thousands of tens of thousands of vehicles in a city becomes an interesting challenge now it may be that there will be some transformative breakthrough in self-driving cars and that's the easy part and the routing is hard it may well be that the routing is the hard part over time. But it's just kind of worse sort of thinking, you know, yes, the not hitting things is hard, but what is it that's going to fundamentally shape what's going on in a city? It's probably going to be the routing, the apps, the kind of

that broader system. Of course, this is why maps, one of the reasons why maps have been become so kind of strategic. And this is also, we talked about this in the last pod, but those three German car companies just spent a huge amount of money on. Yeah. Nokia had spent on. They've got a lot of money on maps. And one layer to that is that having maps now is a kind of strategic asset where it used to be kind of an accessory in the way that your CD player was an accessory and you would out happily have a Bosch CD player in your car and you didn't really mind that they weren't making a CD player.

Today, it's more important in the future. If the map is actually telling the car where to go, then it becomes much more important. Because clearly there's all kinds of interim steps here. It's a bit like steamships that have miles and sails for the first 20 years, just in case. And so, you know, the first self-driving car is not going to be part of a mesh of every car in the city.

You're going to get into the car and say, I want to go here and it will have to work out how to get there. Right. Right. And so it will have to have its own map and its own route. Yeah, 50 years in the future, it may not, to the extent that anyone has the faintest idea what's going to be happening in 50 years time. happens. If we shift from sort of circuits to packets, if we shift to electric cars,

if we shift to this sort of software-driven thing, does it get on the same kind of refresh? Do we replace our cars every two years? Do we, you know, get upgrades every month? I mean, does it become like our smartphones, do you think? Well, you know, someone posted a picture of their Tesla dashboard saying, do you want to upgrade to ludicrous speed for $10,000, which is kind of a quite a chunky in-app purchase. You don't really get that in Candy Crush. Right. But it's the same idea, right? Well, it's an interesting sort of thing in here that when a car becomes electric, it becomes

software, and so your upgrades, you know, instead of paying an extra $5,000 for the sports gearbox, there is no gearbox. So, what do you do as a car company? Do you just say, well, we'll just kind of flip a switch somewhere in the software and we'll make it the sports one? And that'll be an extra $5,000. Do they lend

mail you a badge that you can stick on the back to show everybody that you bought the sports one? Is there like a little screen on the back that says, he spent $10,000, It's sports mode now. It's like that app that there was an app in the early days of the app store that cost $5,000 and it just said, I'm rich. Maybe that's what people will do with their cars.

But it's almost like you're overclocking the car. You know, it's like you buy a software application and you have to upgrade it with a purchase, but the stuff is the code is all there. It's the same with the car. Well, okay, so you bring up overclocking and the argument on the other side will be that people love their cars and people love to drive. And people, you know, love the smell and the sound of a internal combustion engine. Some people do.

Okay. Okay. Some people do, and some of those people will still keep a car, maybe. People loved horses, too. People really loved horses. I mean, way more people loved horses than loved cars. Horses are easy to love. That's true.

And yet, there's a sort of interesting analogy here because, of course, the one of the fundamental advantages of cars over horses, essentially, century ago, was that you have to feed a horse all the time. And so just having the horse of itself was costing you money all the time. Whereas, never mind, you know, keeping it warm and groomed and the vet and the blacksmith and everything else, whereas you put a car in a garage and forget about it for six months and it's still there. Right. Right. And there's a sort of an interesting analog here as you go to a self-driving car, well, you don't need a driver anymore. How far out are we? I mean, we hear from Google, we know that there's rumors and kind of a little

tidbits of Apple hiring Tesla engineers, et cetera. But what stands in the way from this picture that you paint? Well, I don't think there's anything structural that stands in the way of Apple making a car, a electric car, that's manually driven. That's just like time and money and effort. Making a self-driving car, we do not yet have the technology to make self-driving cars. Google is working on it. Tesla is probably working on it, for all we know, Uber may well be working on it.

Apple, if they have, will certainly be working on it, how far they've gone, whether they plan for that to be the car, or whether they will make a manually driven one first, I don't know. But there's kind of, as I said at the beginning, there's a kind of multiple set for blocks here. You can have, and you can have an electric car can change quite a lot without being self-driving or on-demand. You know, you can have on-demand services to change a lot without self-driving. But then when you have, and you can have on-demand and self-driving without electric, for that matter, although probably not very much. But when you have all of that in place at the same time, you know,

your self-driving, I mean, back to electric, it's a lot easier for your self-driving car to charge itself with an electric socket than putting gasoline into it. It's a lot easier and a lot safer. You know, I wouldn't be entirely comfortable having a robot arm, putting gasoline into a car. Electricity is kind of okay. Yes, that seems very elaborate, too, to make that happen. It's a little bit safer. So when you put all of these things together, that's really when all of it changes.

So we've talked about how Apple and Google and others, Uber, et cetera, may be getting into this sort of car business themselves, but is it a good business? And why would you want to, really? Well, I think there's several answers to that. One of them is that, so I wrote a piece a while ago saying kind of the smartphone is a new son. And the smartphone is the dominant. It's the biggest thing in the tech industry.

there are close to 2 billion phones sold every year those are converting to smartphones it's three or 400 million dollars in revenue every year and that's much bigger than they making TVs or making PCs or certainly much bigger than making watches or any other kind of technology you can see a technology product bigger than the smart watch or the iPad or anything else will be so the smartphone is kind of the sun that dominates that

but then you think okay well what else is there is there anything else that is as big as a smartphone this is this thing that is going to be owned by 4 billion people on earth and replaced every two years is there's something else that has that kind of site of scale. And there's toothbrushes and there's shoes and then there's cars. So if you then look at cars, you have this industry that has much higher revenue, over a trillion dollars in annual revenue for making and selling cars. But the unit sales, of course, are much, much, much lower

because the average price is tens of thousands of dollars. Much much higher, yeah. Exactly. And so if you add up Mercedes, BMW, Audi, Lexus combined, you get to five or six million units a year. for over $200 billion, which is almost as big, which is kind of the same ballpark as the iPhone, because the iPhone is, I think, $150 billion a year

in the last 12 months. And so it's kind of an interesting one here because on the one hand, you have this enormous industry, much bigger than anything in tech. Coincidentally, mobile network operators revenue is about $1.2 trillion as well. So cars and mobile networks are about the same size, but much bigger than anything in tech.

But on the other hand, if you added up Mercedes and BMW you would get to maybe the same revenue as the iPhone and less profit. Right. I was going to say that I'm sure there are some cars in the auto world that have the margin of an iPhone, but they certainly aren't made by Toyota. No, exactly. That's not where the money is.

So obviously, you know, you have smaller companies like Porsche and so on that make decent profits. And so, you know, mobile phones are, if you're Apple, mobile phones are a great business, so great gross margin business, cars have low gross margins. Of course, nobody other than Apple is making any money at all in the phone business. So it's kind of an imperfect comparison to make. But it's just kind of, it's intriguing to me just to kind of think, well, what other thing is there after phones? Obviously, there'll be another tech product after phones, where that's augmented reality maybe or folding screens or something or, you know, contact lenses.

But what other thing is there that's as big? Because here is tech that just kind of eats up other industries. but you know Google by itself this year probably will have the same revenue as the newspaper industry the same ad revenue as the newspaper industry the entire global advertising industry is only 500 billion dollars wow yeah that puts it in perspective yeah so and half of that's a third of that is TV and a third of that is internet and a third of those kind of other so like there aren't many other things that are kind of as big as phones and and TV and cars are one of them and so when you combine that with all the stuff that I've reasons like

I've said why cars are going to become more of a technology product and are going to change a lot in the next few decades. That's kind of why it becomes interesting. Does it break down along similar lines? I know you've made arguments for why maybe Apple and its kind of rarefied product lineup isn't well suited to an on-demand car world where we don't own our own cars. But I mean, is there going to be an iOS version of the car, this high-end kind of Apple-y, you know, high-margin thing? And then the Android version where, you know, you can pay as little as you want. Well, it's an interesting one. You know, what will the car world look like in, when this transition has happened in, say, 20 or 30 years?

You know, what will be the distribution of luxury, high-end, mid-range, low-end, sports cars, supercars, limousines and so on. What will that look like? What new kinds of vehicle will get created, you know? There will be new vehicles created in the same way that, like, the people carrier has been created in our lifetime, or, you know, the pickup truck became this mass market consumer product in our lifetimes. there will be new types of car that get created. One suspicion is that Apple will like the high end because it tends to like not making compromises on product to hit prices

beyond a certain point. And one suspicion is that Apple won't make a $10,000 car. One would wonder what happens to car prices as we move to electric and as we move to self-driving. And what happens to the price of owning a car? If it's self-driving, it never has an accident. The insurance is a third of the cost. you're only paying for electricity

so you know the kind of the question was what's the high-end car kind of changes because the high-end car might be cheaper but yeah these are all kind of unknown questions as to how this stuff will shake out well I look forward to the day honestly as a person who drives an electric car not a Tesla a fiat

you know I'm looking for battery technology to get better I'm looking for the software to get a hell of a lot better and Lord knows I would like to not drive so you're using Italian software well made in Detroit though yes which is worst yeah I believe me there's the only thing that goes wrong with my car is a software so I welcome the day that you describe and hopefully sooner rather than later Benedict thank you thank you