Self-driving Cars and the Future of the Automotive Sector
And. Welcome to. Thank. You. So. It's great to be back it is very strange to sit in this room you know there's I've come up here for Google recruiting a few times and we're in the on the newer building but, this is literally I remember my first class in this room in 1997. It's, a little nicer don't get me wrong but it definitely feels the same, so, thank you guys for taking the time in a minute I'm gonna ask you what you want to cover in the session what you want to get out of it so just think about that a little bit I have a set, of things to cover don't worry but, I also want to make sure this is really relevant for you for all of you and. I teach a class at NYU and I do the same approach because this is you're paying for your education this, is your time you're paying your money in your time I want to make sure you get something out of it and so I'm we're, gonna talk about digital, struction, and the auto industry in particular but I can go a lot broader than that, by. Way of background on me I went. To Hamilton College I spent, my junior at the London School of Economics and. My wife in college I, then worked at JP Morgan for, three years I got. Married and my wife and I were. Coupled together a talk so we were actually both we, were classmates, is there anyone this year and right, now who are couples, that came in Mary, Oh marriage. But yeah. Okay. Very. Nice okay yep, so we're still married we're doing great we had a great time so it's been 20 it's been 22 years now. So. My wife Liz and I came in and I remember specifically being in this room coming. Out of here. I went to McKenzie so I actually felt the exact opposite. At that Stephanie did with. McKenzie for about three and a half years I spent two years at a late, stage startup, two, years at IBM and, I've been at Google for 11, years, part. Of it was running part of strategy, and operations and, then I ran our media sales to the telecom industry did. A brief stint, in Google, Fiber which we talk about it but if you want at all and then, for the last 15. Months have been leading our industry, efforts, for the imports, in the auto industry so. That's sort of a rough on me just to give you some background I was, going to talk and particular by digital disruption some. Overall principles and then apply it to the auto industry but, other other other things that people want to make sure burning, questions people had that they wanted to make sure we addressed. Sure. I can make sure to work it all and don't worry okay. Sure. Sure. Google, Fiber and the impact of that industry and where it is where Google Fiber is right now fair, enough good question. Generally. The, world, what. All Carter's right around himself that we changed a lot of things. Sure. So I'm actually modifying, a bit a lecture I did at, Stern which, was half of it was a debate actually on when will driverless cars occurs when they encourage you got the involved and then, the second half what will it mean for all the other issues discussion. Because, no one knows the answers but it's a good set of question. Totally. Okay. Great. Okay. Okay. Okay. Yep. Very. Good question. I'm. A curious about how, Google deals with innovation.
In. Terms of when. You are looking at a certain, development and. You. Look. Upon the market how other tech possible. Competitive. Competitive. Technologies. Are also developing. I also have in my mind the example, of the airship, in the, airplane well, where the airship was almost. Reaching, for engineering. Perfection, then came the airplane all. Of these was destroyed, and. Not. Just for. Self-driving. Cars it's like how, is the Google find a lot of fuel innovation in, this context, okay. Car. Ownership versus. On-demand transportation. And specifically. How you think about pricing, between those two. Regarde. Or. Software. Versus. Okay. Okay any, others right. That's a good list okay, grab that really rushes through the first portion of it right to the second which is great. Okay. So, I'll. Talk a little bit about some of the features around a description how we think about it and how I think about it and, then we'll go into the particulars some of the auto industry what's happening right now but, I'll spend most the time I'm gonna go through that in about 5-10 minutes and then we'll spend most of the time it sounds like on the driverless cars part, of it which is good to know. So. It's an overall guiding thought on ditch on any disruption, that goes on which. Is Bill. And Bill Gates interesting for a Google person to come in and give a Microsoft, quote but, less, happens in two years than you think but more happens in ten years than you'd imagined, goes back to it's Google's 20th birthday right if, you'd pictured, Google 20 years ago you, never would have pictured where Google is today right seven products with over a billion users eighty-five, thousand, employees one, hundred and thirty billion dollars, in revenue and growing a twenty percent year on year in sales. Twenty. Years ago I sat in your seats clearly. If I'd thought of it I would have tried to get into Google at that point instead of going to McKinsey that would have been a much better call I. Wouldn't. Actually be here I'd be hanging out somewhere else but. Because. When you sit inside Google on a daily basis you actually have frustrated you're like we're not moving fast enough and, you would actually say a company like Google e BEC well you must clearly move very fast but, if they're sitting in our seat it feels like it's glacial, sometimes but.
Then You have to look back over a longer period of time and you see incredible, change and the changes are very different than you would have actually anticipated. And. So I can I took a year and a half off to I. Took they went for a year and a half to Google Fiber a lot. Of stuff actually changed any one half and it was actually really helpful to see it because, if you see it every day you don't you sort of don't breathe it but, the interesting impact of this is we'll talk about driverless cars the. Debate I had with my class is in ten years we'll, more driverless cars be sold or. Will more regular driver cars be sold and that's not will to be more on the street because it'll probably take at least ten it will take more than ten years to have more driverless cars in the street but how do you think about what are those characteristics, and what would happen and, the answer probably is that what I'm thinking about what we're going to talk about here in ten years will not look like what actually comes through to your airship versus. Airplane analogy. But. You have to ask you figure out what those right questions are and prepare for a future either but. But very likely not much will happen in the next few years. There, are four main considerations, and I'll roll through pretty quickly here but they'll come back as we talk about driverless. Cars. Great. Harvard Business School professor. Focused. In on you have to think about what is the job to be done on any innovation right, not am, i replacing, an existing product but what is the end need that I am that, I am going after and, when you solve that you generally succeed if you build a better mousetrap you, do not succeed if. You look at disruptions. In any market the airship versus the airplane airplane. Got the job done a lot faster I needed the job was I need to be from New York to LA in a period of time not, I want a beautiful scenic route ride and it wants to feel like a train and all of all these other things it, needed to be the job was I need transportation. From here to there and it did it very very well when. The iPhone came, out and. Eliminated. Basically the BlackBerry they, did a totally different job it didn't just build a better mobile phone. They did a totally different function in, the autonomous industry, we're gonna have to make sure that on cars, for instance that it does it, talks about the job and, that will inform a lot of the questions here I think on ownership, versus on demand because. The job is not, owning. A car the job is I need transportation from one spot to another when. I want it where I want it in an affordable price and that's the job to be done the. Second is you need to think about the value chain, and. The best spots to make changes or we are only changing a small portion of the value chain and. You take advantage of all the other areas, so. One of the things going on the auto industry right now is, there something called Android auto and apple carplay I imagine some of you have seen them if you don't already have it already.
Those. Go in date they have to fit in a certain spot in the value chain but they leverage off your phone so, it's actually a relatively seamless way to get you in there right, it's products, you've already used it's ways it's, you, know Pandora, Spotify, its. Maps though. It starts with those products because it's a easy simple thing to do and all you're doing is changing one thing so, any change that goes on you, should think about how am i using the value chain to get advantage out of it and how am I not trying, to sort of go in too many different spots when. You go in and, BMW. Came out about 10 years with run flat tires. They, actually totally failed I think when your professors did a fantastic book. On it, and. Didn't. Work because it changed too much stuff you, need to slot into one and to make the change easy the. Third is there's something called the Perez search cycle which. Is stuff, starts it takes a long time to, get started and then, when it hits an inflection, point it goes from multiple competitors, to just a couple and. Uh where, are you in that cycle and how do you have the funding to last through this how, are you thinking about your business around, how you get to that and lastly. What is the market ready to be disrupted right. Some, markets are some markets don't make sense like, the automotive, dealer model doesn't, make a lot of sense but it's not yet ready to be disrupted. Because. We'll look at some of the other things later on that I don't think it's fully fundamentally, ready to be disrupted like, we, looked at should you go out and eliminate auto dealerships, because I imagine almost everybody's bought a car at some point now and. Going to a dollar auto dealership, is one of the more painful experiences, but if they're at the DMV and paying your taxes but. It's, still sort of a necessary evil until we find the right scenario, so. If we look at those the jobs to be done is what. Are you trying to accomplish you. Don't buy a hammer you buy the ability, to pound in nails right, how are you trying to do that it. Should change the way you think about it if. We think about the value chain, there's. Sort of four main functions of value chain there's content, curation, there's. The device and then there's distribution, if. You're gonna change in driverless cars for instance and goes back to the question on do you do software and/or hardware, if. You try to build the car change. The software build all the programming capabilities, for it what. Are your odds of success word. You find a way to take advantage of the value chain and change, just one or two things in it like. In an ideal world you'd wind up with driverless cars that would be able to work with any existing car that's, probably gonna be hard up interoperability, stuff but. Should you be able to get it so that it plugs into almost any car that's newly manufactured, that. Would take advantage of the entire value chain and make things a lot simpler that, is what Android auto is doing he's saying I'm not gonna build new cars I want, to slot into every car that gets made right. Now you have an something, like Android auto or so carplay already it's, your phone like, how many views weighs on your phone while you're driving. Good even up here it's probably helpful right definitely you get down to New York City extremely helpful um. You're. Just basically by putting Android. Auto or car play in the car you're just taking what was an RD existing behavior and making it easy to change because, you're changing one thing within the value chain. The. Other one is you, actually need to be number one or number two in any almost any market to win. But. We people talk about Facebook, and Google as it digitally. And they talk about it not the nicest ways, but. Underneath it in almost any market there is there are one or two winners, there. Are increasing returns to scale right, this is not a new phenomenon but. It very much exists, in in. The digital world and it, happens faster, in the digital world than it does in other markets right. Google's been around for twenty years ten, years and it was fairly clear that Google had sort of won in search right. What's, interesting is Amazon, now is coming out with Alexa, and may, actually obviate, the need for search right, the, threat is not that someone does search, better than Google the, question is that someone solves, the the thing that you're trying to solve with search better.
That. They don't look and require keywords they don't require you to go in there they don't send you to different pages they, just give you the answers and they anticipate. Your questions and. They help you in your life and they solve an underlying need so that all of a sudden you're just not going to Google 50 times a day you're going 30 and then you're going 20 and then just don't go at all. But. It's, unlikely that someone's just gonna beat out Google what Google did you do see it in some markets occasionally, I would, argue Facebook was not that different than myspace, but, it's, and you see Spotify. Beat out Pandora on Pandora her and early lead but. It's not that common, you usually tend to see stuff. Like you saw in, in. The cell phone market right. So everyone probably, everyone's old enough that they had some point had a razor phone, right, and. You, didn't change up off the res you probably bought one you might have even bought a second one then. You came around and what what came next after the razor. Pardon. Yeah. But it never really took off what was the next big one. Blackberry. And the BlackBerry crushed, the razor like, if you were a motor like Google bought more Motorola, for a while and. It was one of those things have like it was the highest flying and when the razor came out and then they made it different colors they did all these little tweaks but. The BlackBerry, came out solved a fundamentally, different issue it's did he was an email machine and it just happened to have a phone okay. So blackberries at the top of the world who comes next. IPhone. And now blackberries. How many people go on a blackberry now. Okay. The, iPhone has stayed for a while because no one's come out with anything that's fundamentally, different than it right androids, done okay but it's been sort of a comparable product the, next one will have to leapfrog it but the iPhone did stuff that they never even tried to solve with the blackberry right it wasn't just an email machine it was access to the entire internet and then, it got integration. So we could all get all the apps in there did your music you did things that the. BlackBerry. Never tried so, those were innovations, where it leapt and created sort of a new market otherwise these markets tend to be stable but. If you look up here and you look at a lot of these five, for 5-10 years you should expect them to be roughly stable unless a discontinuous, change takes, place I'll. Dig into one other music. If. You go back 10 years right, Pandora, was around but it was relatively small scale was still the radio right, then Sirius XM came out Sirius XM started, to change the market then, Pandora came out and really really, sort of grew share and then, spot if I came in with just a much better service it's, one of the few examples where Spotify, just did it better than Pandora right, across the board and one out in that one but, now how many of you guys have Spotify.
How. Many of you guys have any service besides Spotify. Okay. Couple okay um. But Spotify, is basically one that market as much as there's google play music and other things it's, going to take a discontinuous. Change to win that. The. Perez I. Don't seen this before. So. It's an application of what's called the s-curve, but. It essentially says you're, gonna have an early gestation period, you're. Gonna have sort of a big bang in an eruption this is usually when you get a lot of venture capital funding and then. The things gonna get to a spot where you really actually know what the use case is and it's going to take off. I'll, give you a couple examples Mobile. It, was the year it was supposed I've been at Google 11 years first, year I got there it's gonna be the year of mobile not. Second, year it's gonna be the year of mobile third year it's gonna be their mobile it wasn't til actually I think 2006. Or 2007 when the iPhone came out that it really became the year of mobile and then it went from. I'll. Give you a Google inside language right I remember, being in a meeting and they were like 10% of searches take place in a mobile device and I was like that's, a real market right that's actually interesting and it, went from 10 to 50 in four years right. And. It should just took off, you, see Spotify, right, Spotify, is doing really well now early, on it was a slow. Burn then, it really took off once everybody decided okay I don't need to buy all these things the, idea of ownership over to on-demand so some of the earlier discussions really took off in a, phone scale you. Generally tend to see this what you see is lots of players, what's. Looked at this - Bitcoin right, - the some, of the financial markets you guys are probably looking at there's, dozens of players probably some of them recruiting up here some of them probably evoke mega work afterwards, there's. A hundred probably plus players within, a couple years there will probably be two to three and when. That market really takes off right now it still hasn't sort of found its groove but it will find its groove hey, Steve. Um. And. Then. The last thing is what are the signals how, do you know the markets ready right. If, you think about it in, order for a market to really be ready you need solid economics. And a, lot and a large scale right. To really be a market that's worthwhile to go after you, can see innovation and disruption in smaller markets but it's much less likely because. It takes a lot of energy and upfront work, to do it so you want a large addressable market and, you want have some idea of what your financial models going to be so. To the question of way, Moe and Google on cars. The, model is probably, likely going to be either a sale model or a subscription model it is, unlikely, that way mo will build its own cars right they've already partnered, with Jaguar, and with Chrysler, to build them and a, few other manufacturers, and it's, likely that that will be the model associated with it but they have a good idea it's a massive, addressable market almost everybody in the world drives and. It's got a likely, model that you could sort of see there second. There's got to be clear value to the consumer the consumer needs to get it and immediately see it if, you go back to search.
When. Google was started, to, talk about 20, years ago I think in the second year, tried to sell it for 5 million bucks right. And, no one wanted to buy it because I was like yeah it, just sort of part of the portal page I've got a well I've got MSN this thing is really not gonna be that valuable there. Was clear value to the end consumer though, the Google saw but, the competition. Didn't see and now clearly search, became sort of the core of it and the last is lo trial and subscription, costs right and. It should be relative to your value the more value create the more people are willing to go through some pain to change, over but. They it's. Much easier if you have lower switching costs in. Autonomous, vehicles it's unlikely that the first time you ride in an autonomous, vehicle you, will own it it. Is much more likely that you'll do something like lyft and, get in it and take him back ok that worked right that was good and then, you'll still keep your car and then, it'll probably get to an inflection, point where you're like ok, I've had this car sitting in my garage I'm living in New York City I'm paying five five hundred bucks a month for this thing I never take the thing out insurance, everything, else I should, just get rid of this or in the very least my. Transmission. Went out and I need to buy a new car at this point I will decide not to buy a new car and, that's how you'll likely see the transition go on this but. If you had to first, move to, get us a driverless car buy it to. Be to use it for the first time that's probably too big a switching cost. Ok. There's, also some basically. Cover all these so I'll just skip that over. Now. Switch over to the auto industry I'm going to talk quickly through the near term challenges and then we'll go to driverless cars there. Are a couple key trends going on the auto industry one. Is you're. Seeing the. Emergence of new new models in particularly the top of the autonomous come. Out and crush. The financials, of a lot of the companies in, the last last. Six nine months the. The. Car companies have lost twenty. Five twenty one percent of the value if. You're sitting in Detroit and you work at General Motors you work at Ford you look at we work at Chrysler you're looking ahead and basically the market and all the stock analysts, a lot of probably who came from some talk are saying you, don't have a future right and that is actually impacting their decisions now because, when they look forward the Ford CEO got fired not because they weren't selling cars today but, because he didn't have a plan for the future and so that required, them to start to think about things fundamentally differently, when.
You Have that kind of impact in your board fires your CEO, you'll, start to see changes and. Does anyone know what Ford's decided to do in the last six months it. Relates to the second point of the page. They're. Good. Yep. They're gonna leave the car market, they. Were the car market right. They had an amazing set, they're. Gonna move over the second trend is. Now. 60 percent of all sales are trucks and SUVs 32 percent of our cars five, years ago almost, exactly flipped, and. How long it takes to launch a new vehicle. About. Five years. Because. It's there's a lot of technology going into running, one of these things Tesla does a little faster I'm about to get one moment you see if it works really well. But. It takes about that long so you've got a fundamental, shift in the industry that, is helping all the US manufacturers, because. They're the ones who make most of the SUVs and, almost the entire truck market is dominated by US manufacturers, but. That's, gonna wash through as you've got you know everybody else starting to launch all these areas so you're seeing those as the near-term shifts as the. Imports, like a Nissan, like a Toyota like a Honda, step, in or even a Jaguar step in and sort of go after this market you're gonna see some a lot of SUVs, get launched and you're gonna start to see a lot of cars get deprecated, Ford, has been the only one who's bold enough to say I'm just gonna stop making them I'm gonna make the Mustang I'm, gonna make like one or two others that are sort of iconic but I'm gonna get rid of the entire rest of the market which. Is an interesting view because part of the view was SUVs, came around because gas was cheap and now they're saying really regardless, of gas prices this. Is consumers, shifted and it's probably done. The. Other thing you're seeing is and not seeing is e-commerce adoption. So. This, and it's really hard to read up here but, it's actually the question is what. Is the shopping preference. Not how many cars actually get sold this way but, which would you rather buy and, if you start at the bottom most. People like I'd like to buy my books online like other than going to the Dartmouth bookstore which I still love to do and I'm gonna go there tomorrow. I'd, later today right. I don't really go to a book store toys, and games has been decimated, used to be tor shops in every town in America I live in New Canaan Connecticut that, thing's on its last legs unfortunately, as, you work up though automotive. People are still hesitant to buy a car without. Actually, getting into it touching, you're driving it and that's, Eve with a few statistics, there. Are twice as many people, watching test drives on YouTube as there are people doing test drives so. For every test drive in mark that someone takes any dealership you've, watched two or three online just. To get a feel for the car but you're still gonna go drive on the car before you do it and most. People will research four to five cars or vehicles, but. Only visit one to two dealerships but they. Still feel like I need to go to the dealership I'm gonna spend forty thousand dollars on this thing I want. To be there and discuss and negotiate into everything else so it's an industry where you'd say there. Are 17,000. Dealers why aren't we having consolidation.
Well, The consumer still looks down at it and says I actually want to go through that this, is painful as it is but, I also wanted to go she hit the price I want to see the thing I want to trust a person that I'm gonna get serviced in those sorts of things so it's. Pretty surprising me when I saw this because I got in there thinking no one's gonna want to buy cars this way comes. Out its regional no surprise in LA twelve percent of cars are actually twelve to fifteen percent of cars actually, get sold and delivered with someone ever never dealing going to the dealership New. York is going to get to the same direction you. Get to Cleveland, probably, going to take a little longer you get to more rural areas probably gonna take a whole lot longer so some natural adoption curves of the, two coasts, in particular, New York LA LA. San, Cisco tend. To lead no, different than Hoover and lyft adoption, and then the rest of the nation sort of catches on. I'm gonna process like the bother you simplifying, Geass is an action by our Almighty. I think. It'll have part of it the, question, will be people. Like to negotiate as, much as we've been attuned to negotiating. I think, that'll have some impact the other part of it is you want to know there's someone's gonna take care of your car so I think they'll have an impact, the, interesting question mean I was talking to an entrepreneur about doing. Something in the auto space is. How. Long will carvanha be relevant, right, the, the dealer could, go away and this is being recorded if, if, hypothetically speaking, the dealer were to go away I don't think would be because the market, all, the sudden said hey I want to buy a cars an endpoint I think would because people stopped, buying cars period or, because, you had a concern a significant, consolidation, in OEMs and less differentiation, and, it became a much simpler more Trent transparent. Transaction, right. Now there's so much complexity, in how you buy a car that it actually is somewhat helpful to go there you'd, have to reduce a lot of that complexity, um. But, the most likely thing actually I think at the end and I'll sort, of put my view out there is I think it goes subscription, much more than it goes purchase. How. Much do you think of that behavior, is driven. By actual. Consumer, preference, versus like blah. It. Could be it can be part of that definitely. Can be part of that in the as a result the OEMs, the the, automakers, are not incentivized, to actually build the e-commerce capabilities right. Because they're not allowed to sell cars they. Have to go through the dealerships it's sort of cheap when it's the old model and, so therefore most. Of the e-commerce stuff needs to be built by individual, dealers where the economics, are not quite as compelling so that definitely could be part of it the, question was what would cause that to change you. Could see so we there's about there. Was about 10 dealers that owned at least 100 hundred, plus dealer or 10 dealer groups they don't at least a hundred dealers. They. Could step in and start to change this the. OEMs have a tough time changing it in their launching stuff in the UK where they don't have these issues and in Europe does that complete over but. If you can't sell the cars and the OEMs can't be there wants to do it it's got to be the dealers that actually start to do this you, are seeing some examples in New York there's one Honda dealer in New York who's out in Queens who will, actually do everything for you remotely you never have to go the place so. You could see some steps in that direction definitely, some. Of the complexity still seems to factor in for folks, but. It's a it's a very fair question is it you, know what is causing, it and what is keeping, it from actually happening. So. That's. Most. Of what's going on the industry what I'd love to talk about that is self-driving cars and sort of when we see this technology happening, and. I'd actually love to if it's okay have a little bit debate with everybody.
So. The debate would be what. Needs to happen in the next ten years for. Actually it to switch over and I'll, give you a couple facts to sort of consider on this and I'm gonna make the argument that in ten years there'll. Be more driverless cars sold that year then, there will be sold regular cars not on the road but but but lease car sold and actually have this exact debate with digital TV about seven years ago in this room and I said it would happen in Google would would, drive digital TV and the, class was entirely right and I was totally wrong so it's, to, some degree take a question of how long do these things tend, to happen in, that case digital, TV wound up not being what we expected it took ten years rather than two but. I want it being different than what we expected so, the debate here is, one. Can anyone tell me how many cars are in what percentage of cars are used at any given moment. Focus. On okay at, night it gets that during the day it's about seven it's about 15 to 20 percent of cars used at any given time so. What you've got is 75. To 80% of your capacity sitting. Idle most. Of you probably drove here a lot, of second years live, off-campus your cars have been sitting there all day you, probably drive twice, a day and then probably a couple of times there later on so. There's a real question of your car sits there it is an asset that you have paid a lot of money for and generates, no returns throughout the day now, you've seen some services, that allow you to now rent your car while you're not using it it's, a little bit kludgy, if. All of you could use lyft and it could get you here and take you back home and you knew you could get it whenever you wanted it one wanted, it and you could do it for three to five dollars every time you do it would. You rather do that or would you rather own a car. So, for me it depends so for example on the environment card and, provide the additional video, for example if I know. This my car I can store something. Ok. Ok. I also enjoy, the value, of a spontaneous trip. Somewhere further and, that. Seems to be very expensive. Okay okay. Nothing. On the car, I mean having having, everything. Taken care of me and I feel with service, I don't, have to deal with the stress of worrying if my car's gonna get broken into or paying. For, parking. Etc, so. Wandamae, an option would be my apartment okay. So. Did anyone see Reid merrymakers, well probably mostly red merrymakers, latest report but it is 200 page in news it was one page of like hundred and fifty-six um. It. Is actually cheaper in five cities and in the United States already choose, uber and lyft for everything you want to do than it is to own a car the. Economics, roughly breakeven in New York of owning a car is about gift about fifteen thousand miles a year right. When, you go to driverless, if, you if you took the belief that we can do driverless with subscription, you've, eliminated a, lot of them some of the maintenance costs but you've been nominated the driver the. Cost should probably be half so. That will say that anyone who drives less than torn in New York City twenty. To twenty five thousand, miles should, actually not on a car but. It'll also change the break-even economics, in a lot of other cities assuming. You had broad availability of, these air services. That's. What, is the sweeper, I feel. Like in, a rural area yep, the dynamics. The economics, change for, us they do it, will be the last to adopt right. Because you won't have the you. Won't have the instantaneous, access to it right in New York City you can get an uber probably within three to five minutes and in San Francisco definitely when, you get to rural Ohio right, it's gonna be harder you're never gonna get to those if you live on 40, acres right, to get to those it's gonna take a while to get to that spot it, will take much time, and. Yes. Although the trip would be the exact same for, whether you whether you're driving or whether you're in it. It. Will be there is a real question of people some, people go up the Bubba effect - which is like I own, a car this is what I do I'm not going to get into some other thing like that and, also, from an economic standpoint the economics, in less densely populated areas, with much tougher it will be much slower to adopt there much the same way as uber and lyft have done it hates. It but in general you see technology adoption sort of going that. There's. A very good point there's a lot of argument that says you have to figure out what is the first or a second thick use case what, is the first use case that people will do for it, people. Are older who have trouble, who. Have trouble driving, right, at night in particular um, for. Me like you know even if I don't get rid of my car I don't, drive at night to go to dinner all that often anymore it's just a few have even two drinks why risk it and, then you've got a lot of folks or handicapped and have issue other issues too that the, other one is the long-haul truck driving industry is.
A Very expensive challenging. Industry that might be at a use case but you're right you're gonna work to use cases and there's, gonna be use cases like world that are gonna take a lot longer to work through. So the question I would ask is what. Would make parents, be. Okay with not having their child learn how to drive because now you have a whole new generation of, people, who actually. How. Many people grew up in New York City or any major city, area. Okay. We're. In Manhattan just. South side okay, how far outside. Half-hour. Okay so yep so you had Drive yeah but, if you look at the people who go to college who grew up in Manhattan there's a fair number they go without a building. To drive um, we, want to put that question it's sort of like Dartmouth instituted, a swim test like. Should you should. Not Mimi why not require that but you would get that question of should, you at some point as a life skill learn how to drive in the very least I do, know some people who have older who don't have a driver's license at all but. That's a very fair question. What. Else would keep it from adapting. We're getting in New York City already exist yep keep, out places like LA like trying to build more tunnels to make that work. Why. Do you need an. Individual. Driving. Car are individual, cars running around use. Netball. La. La what's. That did you drop any I lived in la expand, it's terrible the translation is terrible although the one trying, to. But. Yeah I think. Yeah. Okay. And I would also sorry to New York City I think you're totally right like you there's enough buses enough subways and everything else like that I go, to LA probably once every six weeks the. Industry that first got hit by this was, the car rental industry I will. Not rent a car when I go to LA it's just not worth it right I take uber and lyft everywhere I go I would love to say that I figured out the public, transportation system and done. That um, it's, fair you will in some spots have such a good public infrastructure that, the question would be is it worth it right if I have a known route and I'm using that known route over and over again then yes I would not I would not do that. Personal. Transportation. Not. Like. You can definitely see that there's actually a question whether it will actually go straight to draw. Okay. They're, still working on the redundancy. Come here are you. Anyone, from Detroit might be trying to go to the Upper Peninsula they're, not even giving state boundaries it's. Not it could be in a state local ad student and they can be going from this and doing this to ice powers. Um. So, yes. I think there's some very fair questions if we go through some of these areas these, have. They covered all of the use cases like I would not get on a driverless car like a Tesla and actually taking around this area in the snow we were having this discussion this morning like that would terrify me, like. Weimer was mainly testing in Arizona in California very. Nice climates, very easy good visibility yes. It will take some time to get through that um, the question of that is once a being machine learning and how fast you can actually get the technology and how fast you can get the inputs right, now I think Google's at eight million miles driven autonomously. With. Someone either as a secondary or even without a secondary, and every, time the moment the more the, basic view is the more in more, data you get the better your machine learning algorithms will get in the better to be able to anticipate those things and so, the Michigan one where you're going from Detroit to the Upper Peninsula through, multiple, climates and through, different types of roadways highways, and other, areas you'll get such great you'll get so much learning on it that it'll actually be able to cover all those but, it will take time because, the early users won't have quite as good of that and it will only be as it builds.
Yeah. You, tell it yep. Um fair question right like you know the, I drove up here yesterday there was a very good chance my car wasn't gonna make it I'm not very good results. In my new car Tesla, is the same issue with is I do I have enough charging stations there, are issues like that that will come up the, questions on those ones will be are. You to have some system in place they can sort of accommodate, for that but, if the lidar goes do you have the ability to just back up into from. Cat you know the levels clearly from level 5 autonomy, which is the top level autonomy back to for right, now Google's built way Mo's building their cars with no truck with no with, no steering wheels so, you would have no backup the way to do that so that's a definite risk. Because there's. A bunch, of problems including, self-driving. Car train country works go super slow because it was beautiful night -, yeah. So. That's, one of the things like. Yeah. We're, that, so. In an Arizona apparently, they're testing all these vehicles and they had put a bunch of people adopt in who just are using autonomous vehicles and people. Are getting ignored cuz they drive so slow and they stopped so frequently that they're like going around them at stoplights and we'll, the vehicles like I got, to make sure I don't hit anybody right, like they're the safe like if you hit somebody it's a massive story if you go slow it's just. So. One, you prove good question you I would assume you would have to give up that but on the flip side abut you to be able to get anonymity. Right. Just like you, personalized and probably log into Google or log into YouTube but. When we go to use all the data for you we hash you into something that so we can't back a to somebody else and so, you would essentially get all the data and you would know this person, with this hashing, did all these things but, I would not know that's you by name so. That's where the, general, data privacy restrictions in other areas that are going on are gonna have a real impact on this interesting. Lee sort of goes and then also in the a blob of security right anyone watch the past imperious, a right like could somebody AK into these things and take them all over there's. A there's a nuclear type thing that you worry about. These. Are all really good questions the other ones that I would think of is, there's. A big I'll. Come back to Google Fiber. The. Government, policy right, it's all state there you have to do national state and local when, you have to make a lot of changes and governments, are not the fastest thing to move that's, why I've seen certain certain. States go much faster California. Arizona you'll. Probably see a couple other states who will see the advantage of it Michigan will probably be the next one to go because. If you're the big three and you're like we need one of these guys to win otherwise our economies and some trouble so. You will see some government policy but it's gonna be a real challenge to get through the whole thing in particular. You get one bad accident, right like Tesla has had a couple of them and it gets massively highlighted, and the, statistic, that like purge, per mile driven, you. Have a lot less issues with the driverless car statistic. Gets lost in the story right 10,000. People are statistic, one person to story and the, one person the story can overcome a lot of that in particular you, might get a government official who's not willing to take that risk um. Infrastructure. I believe. The belief is you're gonna have to work on the existing roads like changing, the roads would just be way too hard. Nico's. Ultimate, each other I feel like that pushes, the time. That self-driving, cars could calm weather actually quite a far away down the road I didn't see a good answer not. Fair. Although in the tech value chain idea is if you required all the cars to be able to talk to each other you, would then need to change four hundred and fifty, million you know like what. Is it hundred and fifty to three hundred million cars that's. Kind of hard so, I think you're likely to see autonomous, cars and regular cars out there will, you see some bad, behavior because of that to the extent that it happens right, sure once you know it's an autonomous car you know if you shut if you cut them off that you know they're gonna stop people, might just are cutting them off right um, they might start going on they might do some things like that but, I can't see a world where you would go straight from no, autonomy to full autonomy that's, just too hard a change to go through, what.
Will Be interesting to see is to you start to see people just naturally do it themselves, one, of the biggest excelence, to the iPhone was, that corporate, execs when they got the iPhone they, went back into their IT department, in there and they said had I've got this iPhone I need to use it on my service in the IT people won yeah, that doesn't work it's not secure but Bob all you have to stick to your blackberry and, the CEO said yeah, yeah you're not hearing me I need to be able to use my iPhone, and then. They switched, all the corporate infrastructure, to do that so you start to see as you get a tipping point some of those changes actually do occur but it takes a long time to get there and that was not the first used case. By. Versus, build. This. Is the current. Number. Of folks working in connected cars. 515. Companies, clearly. It is not in the middle of the prize search cycle yet. All. Of them have some kind of stake in it if, you look at it the venture capital, much the same way that conventional capital's going after the. Cryptocurrencies. Bitcoins. Etc, you're, seeing massive investments, here the, question is who's gonna win we, know from some of the earlier discussions there's gonna be one or two winners right there can't be 515. Companies but, just the acquisitions, you'll see companies, get merged and you'll see new, ideas emerge that will make this better and better but. The other thing you see is 515. Companies that want to do this well, lobby they. Will push for this right and they will create a, ecosystem. Because they have all have something to gain will. Really change in the last two to three years is GM really went after this now, Toyota's going after it all, those people were lobbying against, driverless, cars because it hurts their business now, if they think they can win they will draw a lobby for driverless cars, so. You get your buy versus build you're. Gonna sit down here as a company and Google has done this and other way, most on this and others will and say what, is it I uniquely, need to do and then, where is it that I can leverage and. I'm not man if I don't want to build everything myself so. Way mo decided not to build the cars right to help design it someone else does actually all the manufacturing, there's, components, in the car that they they, you know all, the electric stuff I want someone else to do i strictly, want to do the software the lidar and the, driving and making, sure that it drives within. Safely. Right and within and, get you to the right space right time I don't want to do the tires I want I don't want to do the acceleration, I don't want to do all that other stuff so that's, buy versus build liability.
Of Self-driving massive. Open issue you, guys are pointing out there's a lot of massive open issues in this and all of these could keep this from happening in ten years right, when, a driverless, car gets in a crash so the test the ones when there was in self-driving mode and, the person got in an accident who, pays for that right, now you're choosing to put it in self-driving mode and I think they're making the end consumer do that but, if you have no steering wheel how can you be held accountable for an accident you couldn't you couldn't do anything so someone's, gonna have to step in and say they're gonna pay for that he. Hates that but that's largely more likely a large company right a small company can't take that risk one accident could take them out right. But insurance, is clearly gonna need to emerge, the. Google innovation next waves the impact of the winners. We're. Doing a lot around the automotive industry there's, Android auto, there's. A lot of things we do from a media standpoint they buy three three to five billion dollars globally, they buy five billion dollars in advertising from, us right, and we're about to go into competition with those clients right. That's create, some innovation and some checks and balances the interesting challenge I think people have gotten better and better about it is everybody. Knows they're gonna compete they're gonna collaborate in some spaces and compete in other spaces, in to the extent that you just say tain your swimlane, you're, probably long-term potential is not that great and so that people have gotten just more and more comfortable with that. In, terms of buying software or the it. Makes it necessary powders so. Let me advertising, in the cars or is it going to be some other oh, yeah. I don't yeah I think we're gonna have to figure out we're gonna have to make a model beside the advertising, work um, I think, advertising at this point is 85, to 90 percent of Google's revenue which. Yes you'd say okay we're predominantly, that the, cloud, business is done okay a couple, of businesses have done okay when. We say it's 85 to 90 percent that actually tells you that 15. To 20 billion dollars comes, from outside of advertising, most. Come that would be a decent sized company somewhere else, so. Yes that is an issue though right like the, core things were really good at it's, gonna have nothing to do with the core things like I will, not be touching at all I'd work on the media sales side of it I will not be helping them figure out like how do we monetize that there, might be some advertising unit you might do some portion portions of epital be a small thing relative to be a well service so. Yes there's a challenge for Google and for anybody else to make sure that they can do that model thankfully. Lyft and uber started, to prove that model right, if you want to do it it's a subscription model they started to prove it out. The. Ownership versus on demand I think. The interesting question here is just what is the advantage to own versus. On demand if you know you can get a car so your real question becomes a real very real one but, if you're in an urban area and I know, I can get a car and they'll guarantee me a car within two minutes why. Would I ever want to own one unless. It's a big economic advantage. To me but I probably need to do twenty-five thousand miles a year do. I want to do that now. You'd say well car, ownership has been an important fundamental, right for people for their entire lives but. Look at all the things you now do that are on demand that, my generation when I was growing up or Steve generation didn't. Do right. Like you. Know the idea that you would go Airbnb. And stay somewhere and not own it right the idea that you know even renting, versus owning was such a big big, big change people, already made that decision right, Rent, the Runway came, in and created a whole new market right those things I, think, a lot of those there, will be those issues the question is will you get enough adoption without it and.
Then So lastly software versus hardware to, the earlier, point on. The whole. Frame if. You want to do the entire thing you're probably hurting yourself, right the question will be how can you interoperate, with what is already out there how, can you make this work in most cars not just five or ten or, how can you get those five or ten to pump, enough volume that they could really work so. I. Know we're gonna finish up here what I would say is I don't know what the answer is like the Google TV thing I sort of Adam and they argued for and I was wrong um, but. I would go back to the Bill Gates quote right, not. Less will happen in the next two years there will not be a ton Google will go from or way Moe difficulty will go from eight million miles on the road chip probably forty to fifty million if not 100 million roads and get phenomenally. Better but. This is the kind of thing we need to be perfect it's not like search where you're like oh I got a bad link I'll just go back you. Know the, something goes wrong it's a bigger issue, but. If you look at the value that can create it on this and you go back to the jobs to be done and you say I need to get from place a to place B and I don't want to spend that, incredibly, much money if you actually look at how much you spend even sitting here in this room on your cars, on a monthly basis if you could take that down by half and get the same quality of service would. You be willing to do that and not worry about DWIs, not. Worry about like taking your kids to school not worry about your my mother is 82, and she still drives our own car and. Gives. Me anxiety, um, right. First does not worry about those things if, you can fit within the existing infrastructure. And make, it so that it's a smaller change, right for individuals, right so, that the first time you do a car it's not I'm gonna go switch over and just go whole hog and buy this thing instead, I'm going to increment it the same way I did it over and left and then when a big decision comes up then I make that decision right. I don't switch from I'm not gonna have a driverless car to I'm. Gonna have it I'm only going to be self-driving you're, gonna go and take it a few times in probably major cities because there'll be a service you'll, keep on that service for a little while then, you'll start to get to the point where you're like okay I'm using it more and more and it comes to your area and then, you finally get to the spot where like I need, to buy a new car do, I want to spend $50,000. To, do that or am I better off just actually relying on this service and so you'll have a slow road to do that that. Is sort of the question I would argue that some of the phases can happen but, I would also then say say that's the current path ahead of what you can see and. I'll go back to Bill Gates lest happened into but more happened in 10 and. Say that's probably actually me bigger, I'm actually anticipating, you could see the drones you could see other models emerge but. I think it is I personally, think it is highly likely that our children that okay my children will drive my kids are 10 and 15 they're definitely gonna drive um, and. I won't let them go, off to college without actually, having learned to drive nor, will they let themselves um, but.
I Think it's likely that children. Of the, people in high school or even your generation, you know that's 20 it could be 20 years from now will. They drive predominantly, or will they even just say I don't even need to do it at all right, nothing real, question thanks. For the debate there's a lot of fun.