 So, let me quickly introduce the panel and then we'll dive right in. Villaleta Bolch is the European Commissioner for Transport in Brussels. Carlos Ghosn is the CEO of the Renaud-Nissan Alliance. Paul Jacobs, Executive Chairman of Qualcomm. And Wendell Wallach, a scholar at the Interdisciplinary Center for Bioethics at Yale, who spent a lot of time working on these issues. And Carlos Ghosn, I'd like to start with you to just tell us where we are with the technology and the adoption and what the roadmap looks like for the next five to ten years. Yeah, well first we need to make a distinction between autonomous cars and driverless cars. Okay, there are two different things and there is a lot of confusion in the articles. People think autonomous are driverless, it's not. Autonomous car, the driver is in the car and he decides whenever he wants to drive, whenever he doesn't want to drive. Okay, so it's about empowering the driver. You give him the opportunity to say, you can do whatever you want. Okay, this is now, this is coming now. And just to give you some steps, one lane highway autonomy is already there. It's offered on a lot of cars. I can give you the example of Nissan. It's on a minivan sold in Japan today. 60% of the people are taking this option and they are paying for this option. So it's strong. There is a demand from the consumer. 2018, you're going to go on multi-lane highway autonomy. 2020, you're going to have an autonomy city driving, which is the most complicated case. Okay, then you have the driverless cars. Driverless cars, a car without a driver. Obviously commercial use, robot taxi, Uber, et cetera. There is a lot of interest on this. And here it's going to take a little bit more time. And here we're talking about more 2021, 2022. Now, I'm talking about mass marketing. And I'm not talking about having a prototype working somewhere or having a fleet of 500 cars. This is already there. I have driven, in Palo Alto, two weeks ago, a city driving autonomous car, an Nissan car. We have a prototype that's working very well. But we know that between the prototype working well and the mass marketing, it's going to take two to three years. Because first you need the thing to work 100%. You need to take in consideration a lot of cases where it's difficult to make a decision. And also you need to wait for the regulator to allow you to put this kind of functions into the car. Okay, good. We'll get into that. I just want to emphasize one thing you said, which is the consumers, using the Tokyo example, you have a minivan on the market that can drive in a single lane without the driver interaction. Consumers really want it. The World Economic Forum did a study saying that people are willing to pay as much as $5,000 extra. That your experience? Yeah, it's our experience. First, he's very interested. We have a lot of feedbacks about this technology. But the first reaction is comfort and less stress, okay? Well, until you experiment this technology, you don't see the difference between driving with your hands on the wheel, eyes on the road, paying attention to everything, and driving without having to intervene into the car. There's a big difference in terms of stress, particularly, let's not forget, people spend, in average, on the planet two hours in the car a day, two hours a day. In the United States, it's a little bit less, it's one hour. But this is an important element. On top of this, you need to take in consideration the fact that it's much safer. 90% of the accidents on the road are due to human error. The more you put autonomous drive, the more you have driverless cars, the more you're gonna reduce accidents. Paul, the connectivity and the infrastructure to support this is part of what you do at Qualcomm. Do you see the timetable about the same way? Yeah, well, I mean, we're, so the industry's working on 5G, and 5G's all about mission critical, high reliability, very low latency. Latency means like I do something and when I get the response. And you can imagine that that's going to be very important if the cars are talking to each other and there's a car up ahead that stops and they're following each other, kind of platooning for aerodynamic reasons. You don't want the one to stop and the other one not to find out about it too late, right? So that's one of the things that's coming and that's 2020 timeframe. Right now, we're looking at platforms that come out of the smartphone space. We've got machine intelligence in there. We've got computer vision in there. We've got all of this sensor integration. And we're doing things where when we allow that kind of capability in the car to talk to not just the cloud, because that's a little slower to go to the cloud, but to infrastructure in the city, I think that's where we're going to really see some benefits relative to deaths. I mean, there's 13 pedestrian casualties a day in the United States. And we're going to fix those kinds of things as well. It's not necessarily autonomy, but it's in the car talking to the smartphone of the people that are around, because the car knows that it's coming up too fast to a red light or something like that. As you know, there's a phrase in Silicon Valley about the minimally viable product. You put something out there quickly and then you iterate. That's not a good approach for this technology. Personally, I hate that idea. I mean, I think the whole notion of beta testing, mission critical software on your customers is probably not a good one. And we've even seen that happen. After the Tesla decapitation accident, the guy gets up and says, oh, well the feature was coming in the next release that allowed the car to detect a truck coming across. In my mind, that's not a product. So we can't do that as an industry. I've already been through a number of cases where technology got rolled out and it disappointed the customers in one way or another. Handwriting recognition is a perfect example. If you go way back when, they had handwriting recognition and an Apple Newton and everybody mocked it. And we didn't have handwriting recognition. We still kind of don't use it, right? So we need to be sure that we don't ruin the customer's expectation and set the whole industry back 10 years by doing that as well. The regulators ready for this? Well, that's a very good question because there is a lot of things happening at the same time. Mobility is being pretty shaken up with many different trends. First of all, decarbonization. We have a lot of changes now happening because we really wanna decrease all the impact on the nature and transport is one of the biggest polluters and especially road transport. So here it's a big push for alternative sources of energy and new propulsion systems, new technologies that will really enable this. So this is one track. The other track is the digitalization is now offering all this incredible functionalities that not only make the drive more fun, but as colleagues already pointed out, really addresses the safety issues which is related to connected vehicles, cooperative vehicles, autonomous vehicles, and then of course, driverless vehicles. So there's lots of concepts. At the end, somebody has to pay for it. So there has to be a consumer who is willing to pay into all these changes. So I'm a computer engineer by profession. I'm very enthusiastic about what this is happening, but we really need to understand the human behavior and the statistics also show that people change the car seven to 10 years cycle. So if we take that into account, I don't know how these figures that we've heard from the industry could be really met, but of course, there is no doubt that this is the trend, that we are going in this direction. On the European Commission level, we changed the process that we prepared the regulations and we are now in the early innovation cycles together with the industry. We co-create and understand what the technology, how the technology will be driving the changes on the market, so we try to tap into this. But I will be a bit cautious about the speed of adoption. The speed of adoption because of the adoption capacity of consumers, maybe 1%, 10%, but let's talk about 90% people in India, people in China, people in rural areas. It's gonna be an interesting trend, but I'm sure that this change will be happening from different sites. So we're gonna have many roads that will lead to the final destination. We know already today, I fully support the idea that autonomous cars are already here. We see them in controlled areas and campuses, airports are using this technology to move freight around and things like that. And this will be just getting stronger, stronger, stronger. But to bet on the mass introduction of this, we will probably be a bit longer time. You're talking about consumer adoption. Yes. But what's the regulator's view? I mean, I just put up this slide that shows many of the public benefits. You already talked about pollution, more efficient spending. So there are a lot of big gains to come from this. Do you feel you need to slow the process? No, we're not gonna slow the process. As I said, we are working in the industry now. We are reshaping the way how regulation is done. But we are staying technology neutral, which means we put out, for example, last year, two important strategies, one on CITS, which is cooperative intelligent transport, and the other one, strategy on decarbonization. We're very clearly stated, what is the direction and what is the final goal? But how the technical solutions will then emerge within these strategic orientations is, of course, up to the industry and up to the market to decide. Wendell, all sorts of interesting societal issues, ethical issues, you know, insurance issues, how you, what, how does the technology answer, what do you see as the sort of key sticking points here that have to be worked through before we can move to fully driverless? There's truly a broad array of them. I think the first one was who's truly accountable when these systems fail, particularly if there's a trade-off between the car and the driver at times. But that's been largely solved, because I think the industry has recognized that few courts are going to hold the driver or the human driver accountable if something goes wrong, and the industry has more or less decided that the benefits are there economically where they can take on the accountability. The big ethical challenge that people have been talking about for a long time, which is sort of a false challenge, is this trolley car problem. Trolley car problems have been around since 1967. They're really ethical thought experiments about what people would do if you have to make a choice between saving five lives at the expense of another life, but it requires an action on your part. And these have recently been translated to driverless cars, and there's been a lot of publicity around it, and there's been some research on it, and it's found that consumers in general say the car should kill the least amount of people, even if it means it kills the driver or the occupant, or at least the humans who are occupying that car, except nobody's gonna buy that car. All the research shows that. So in order to stop a once in a trillion mile event from happening, we want the car to take an action that will stop millions of people from buying driverless cars, and if they don't buy driverless cars, we're likely to see thousands of deaths that would have happened that perhaps could have been stopped with the use of driverless cars. It's just showing it's kind of a false challenge, but the interesting thing about this particular challenge is it's brought to our attention that driving is not simply a lawful activity where a car stops at a stop sign, and if it sees a ball, it gets prepared to break in case a child is anywhere in the vicinity. No, it's a social practice, and there are these countless different situations that can arise that there is no simple right or wrong or perhaps even technological solution for what the cars should do. So consider four cars coming to the four-way stop at the same time. Who goes? What happens if the human drivers are gaming the driverless cars? Consider New York City and the taxis trying to gain the driverless cars. The driverless car might be stuck at an intersection for a long period of time. So we have this array of both practical and ethical and finally legal and accountability issues that are coming into play in this space, and we don't have the technology yet. The technology isn't ready for prime time to solve them, nor have we as a society decided how we want some of these challenges to be handled. In other words, they don't get decided by existing efforts. We have to come up with new norms, we have to agree upon what we would like. Carlos, how are you dealing with those questions? You have to... It's totally logical, I totally agree. That's one of the reasons I'm telling you. Autonomous cars are immediate. Driverless cars are going to take some time. Not only because of technology, we still have some problems to solve, but I'm very optimistic about the solution to the technical problem. But there are many, many questions which have been outlined here which are totally true about you're putting a machine without anybody in this machine on the roads, and we know that there are some cases where it cannot make a decision. It cannot make a decision. Or if you wanted to make a decision, you're gonna have to make some choices and some priorities. And the car maker alone cannot make these priorities. He's gonna have to go to the regulator to tell him what are the priorities into these cases. So what we're doing now, in order to solve, again, we are in prototyping here. What you are just testing is when the car is in front of a situation that it cannot solve, whatever powerful is the computer you're putting into the car, because the car is in front of a situation where it's a no, no, no, okay? So it stops. Then it has to go to a human being sitting in a control tower and give him the photo of the situation and the human being tell him, okay, cross this line, okay? Because I know there is no car coming, cross this line. You have plenty of nose. This is the know you're gonna break. It's human being. By the way, it's not very far from what NASA's doing. You can ask yourself today when NASA sends something on March, plenty of computer, silly human being is on Earth and making, in some cases, some decision. We are collaborating with NASA to try to say when the machine or the robot is in front of situation where you have plenty of nose, what do you do? What they do is come back to Earth and we tell you exactly what we do. Now, with the power of computer, this situation is less and less frequent because you're solving more and more problem. But we don't foresee a situation yet where the car can solve all the problems in a way where there is absolutely no ethical problem or value like, that's why I'm saying, autonomous cars are here because you have the driver in the car and at the end of the day, he makes a decision. And when the car has a problem, it gives back the control to the driver. But when there is no driver in the car, situation is a little bit more complicated. Technically, we can solve it, but there is still a lot of things going on. But I think what I heard you say early on, Carlos, is if it's a tough ethical decision, it's not yours, it's hers. It's obvious. It's obvious. At the end of the day, when it comes to, between different evil, what you're gonna choose, it's up to society to decide. So how are you gonna do that? Here, I'd really like to invite us all to accept that this will be a transitional, long, traditional period. We all know that these technologies will be deployed sooner or later, but I would like to bring an attention to a really big difference between the classical experience of digitalization that we've had so far, when we were worried about digital identity, or worried about lost or not lost data, about accumulation of data, privacy of data, all fine. Now, for the first time, we're seriously talking about a consumer product that has no chance to go to 2.2 version to be really good. It has to be good at the beginning, because we're talking about lives. We're no longer talking about lost digital identity. We're talking about lives. And that's why I keep saying we will follow that, but we need to learn, but we need to learn without another exam. So this is gonna be really safe and secure steps that we will be walking. Regulators will follow that. We are here, as I said, we are following this development. We are working with industry very closely. We need to engage NGOs much better than we have so far. And to really start thinking about also more complex issues, our social issues, I'm very much fond of the final results because it's gonna bring incredible benefits to social innovation. We are seeing then bringing back to social life elderly people. And of course, for the European Union, that is a big issue. We're bringing back disabled people in a full-scale social life. They're gonna gain independence again. So these are all very exciting benefits that we will see at the end, but we need to walk this path very cautiously and really be sure that the technology that is deployed is safe, that we can use it, that we really know what is going on. But I'm very hopeful. No doubt about that. It's just that let's not be too hyper about it be very, very responsible. I wanted to go back to the issue about the autonomous driving because even with the driver in the car, that driver is gonna get lulled into a false sense of security. And so when something critical happens, do you really want this driver that's been very distracted, not really aware of the situation around them? So I think for autonomous vehicles, even the early versions to work extremely well, there's going to need to be a lot of situational awareness. The car is actually going to need to be monitoring what's going on around it and start to predict when some kind of a situation may come up. So that it can cause you as a driver to be a little bit more aware and focused on what's going on around it instead of reading my paper, watching a movie or doing whatever else I'm doing. I think that aspect of the human machine interface is going to be an interesting area for all of us to work on for a little while and machine intelligence will help a lot and doing these kinds of things where the car is actually monitoring the state of the driver. I think this is going to be an important technology and people don't talk about that all that much. But there are real problematics in the trade-off between a driver and the system itself. Even if the system anticipates that it's coming upon a situation that it doesn't know what to do, that doesn't mean a human who is napping or picked up a book or is texting is going to be able to respond in time. It's at least in the US is starting to look like the courts are going to rule against the car companies in those situations. Therefore, I think the car companies are now understanding they are going to have to take on some of the accountability and liability or at least through insurance or the pricing of the car and handling the insurance elements of that. But I also want to come back on what is really in the hands of the regulatory authorities and even in the hands of the car manufacturers and to what extent we the citizenry needs to be engaged in this particular challenge because at this point we are talking about life and death. In some cases these are experimental technologies that we're putting on the road. That makes us human subjects in research that we have never consented to be in. And that's one thing if somebody has downloaded software from Tesla and has decided to use that autonomy feature, but it's very different when you consider that someday there's going to be a high profile incident in which a civilian, perhaps a child even, is killed by an autonomous car. But we've had that. Well, we have that, but it wasn't high enough profile. It was a Florida incident. And I'm just saying there will be a high profile incident. I don't know what will set that off. I don't know that it'll be the first or the 10th, but we are moving into a realm where technology is moving into the driver's seat as a primary determinant of humanity's future. And we need to be engaged in what standards, what morals, what norms we want to have set in place, and we need to communicate that to our regulators. And so is the constraint, is the biggest constraint here a technology constraint? Is the biggest constraint a regulatory constraint, or is it a public acceptance political constraint? All of the above. Yeah, well, if I... I was trying to get you to make a choice. Yeah, no, but let me... I don't think the technical constraint is the toughest one. I don't think so. I think we are advancing very fast, not car makers only, but also chip makers, sensors makers. I mean, we know that we're going to be at the point where we're going to be solving a lot of problems. I think the issue is going to be much more, I agree, societal values and regulators. Regulators are going to be in front of a lot of choices. But at the end of the day, they are going to have to decide what's acceptable, what's not acceptable, which is a tough choice, I'm going to tell you why, because today, today, hundreds of thousands of people die in car accidents today. So if we compare this technology to a perfect technology where nobody dies, then it will never see the day. You have to compare it to how much you improve to the situation of today, how much life you can save compared to the situation of today. So what we need to take about, what are the situations that are happening today that we can avoid with this new technology? It doesn't mean, and I agree, we have to be very careful that when the technology come to mass marketing, it's done. It's not in development, by the way. I say, oh, yes, I'm sorry, next year I'm gonna have. Sorry about your head. No, I'm sorry, no, we cannot do that. But at the same time, we need to look at how much accidents we have on the roads today, how much debt we have on the roads today, and how much we can improve with this new technology. Yes, go ahead. The big issue is that you've got a human or actually lots of humans in the loop. If we had a system where it was all the cars were autonomous and they were all talking to each other and they all had full situational awareness, a lot of these problems would go away. I actually think the big issue is we will build technology that really understands the world around it. But then it's, you know, this issue of the New York taxi driver cutting the car off. Well, okay, it might make the car stop at a stop sign for too long. But actually what I'm more worried about is as it's driving along and then the New York taxi drivers figure out, okay, this is how that car reacts. And then the next software update comes and then the car doesn't react that way anymore. Now they've got to adapt to that and what happens when there's an accident there. So I think it's actually the human to the system that's going to be the most difficult part. That goes back to this step-by-step, really gradual progressive introduction of these technologies. And I would like to mention, of course, very important standardization issue that the professor briefly mentioned as well. This is something we are focusing on very much, not only on a country by country level, but also on a global level. We want to reach standardized solutions. So I'm glad that G7 group when the last meeting that we had, we formed a special task group that will now deal with global standardization. We're trying to also work together on the alignment of regulatory frameworks that we do have a common approach to that because if people understand the social framework in one state, this social framework should be the same in the other state or understanding of technologies should not change when you cross the border. So these are a very important issue that also need to be resolved. But I fully support the idea that, of course, right now, we have a challenge on roads with the fatalities and it's unbelievably high. Europe is the leading one and I would challenge the US statistics, but US right now is the best place, European Union, how we deal with that. But we have 70 dead people per day on roads. I was just talking about pedestrians. Aha, okay, sorry. But this is, of course, very motivational figure. That's why I'm glad that the European industry got together and we made an agreement. And by 2019, nine different additional features will be introduced in cars, which will improve road safety. This is a little first step towards cooperative. So a limited example, like the autonomous single lane Japanese minivan. Is that reducing accidents? No, obviously the experience is very recent because we start to sell this car one year ago. And it's one model on the highway in Japan. And as you know, usually in Japan, I follow the rules, not a lot of people, you know? You don't have New York taxi drivers. Because you have to imagine these technologies in worse situation. So you need to go to some emerging market city and test this technology. What's gonna happen when people don't respect the rules? When people don't respect the rule voluntarily or why they don't respect the rules because they are safety. For example, I'm gonna give you an example. There are some cities where at night people don't stop at the red light because they are afraid, okay? So you imagine an autonomous car in a city like this stopping at all the red lights where everybody else is passing. So we have to adapt the technology to the local environment and what are the practices. This is where it's gonna get much more complicated. So we're gonna first test in cities like Tokyo, Paris, New York, et cetera, where, well, things are, I would say rather under control. But when you start to go in cities where things are not always under control, become more complicated. Do you think it's even possible that regulators would say, okay, if this part of a city is going to be autonomous vehicles only and then solve that by using maybe car sharing services so that you would go to some place and you would just be transported around by a car that you don't necessarily own. And therefore the system would be not much more controlled. Is that a conceivable? Well, currently we are working on a hybrid solution. So we do wanna see how we can empower non-digital cars in order to be able to participate in the hybrid traffic structures because it would be very, very challenging from the investment points of view to really start dedicating lines. But I'm pretty sure that in city mobility, urban mobility solutions, they might be zones where they will decide to only use these type of cars. And something similar is already happening with electric cars. It's clear that this is really a stage process we're going through. And Paul, for example, is pointing out perhaps where the future of this will be that we will have maybe only self-driving cars in some regions, we'll have them all communicating with each other. But the infrastructure challenges between here and there are quite daunting. So as Carlos mentioned, we're starting with autonomy, with autonomous features in cars, we're then going towards self-driving cars, we will then probably be going toward the networking of cars so that they are communicating with each other. But those later stages are still a lot of ambiguities about how that might unfold. The difficulty is at every stage we can see benefits. And at every stage we can see saving in life. I don't think there's a disagreement among anybody about the benefits. It's just that there are these risks and how are we going to mitigate those risks? And can we take on some of those risks in a way where we still largely get the benefits? By some estimations, you may have as much as 90% less deaths with self-driving cars. I think those figures are exaggerated for reasons that we don't need to go into here. But we're talking about 1.2 million people dying on the road, simply because an autonomous car can break much more quickly than a human can. An autonomous car is not going to be inattentive. I think the cars will be networked much more quickly because getting the connectivity into the car is actually a relatively simple thing and it doesn't have a lot of these other moral and ethical and human interface issues and even if it's not built into your car and there's a lot of connectivity going into the cars right now, it'll be built into your cell phone and therefore the car will have some way or the person will have some way of being networked with each other. So I actually think that will come quite quickly. And no dead zones, we hope. Well, it's also going to go car to car. Car to car, yeah. This whole vehicle to car, vehicle to infrastructure and so forth. So I think we'll get that kind of situational awareness relatively quickly. I'm sorry, go ahead. No, I just wanted to say that that brings another very important topic on the surface which is cybersecurity. And we are very much looking into that and we're hoping that industry will take a closer look to hybrid communication solutions which we include many layers of security which means including satellites, V-fi and 5G. So not only one will be able probably to address it because it's a really serious issue. And no amount of protection will stop, stop incursions from happening, right? But you know, we as a car maker are responsible to make the menu, put the menu on the table, okay? Connectivity, autonomous, driverless, electric, zero emission, et cetera. We'll put it on the table. And then country by country, market by market, this menu is gonna be decided with the regulator. So it'll vary, it'll vary. Yeah, we may have different solutions in function of the market because we know that there are some benefits that are gonna be driven by consumer demand. Autonomous car is gonna be driven by consumer demand. Driverless cars are gonna be driven by consumer demand because driverless car, we're gonna be selling to a robot taxi company, we're gonna be selling to Uber, we're gonna be selling to this kind of company. But then you have choices that are gonna be driven by the regulator. Nobody wants to buy an electric car because it's electric. No, no, the consumer doesn't care. The regulator by saying, okay, this is a level of emission which is allowed, and by dropping this level of emission, automatically it's gonna be starting to move car makers and the consumers toward electric car. And I'll give you a very simple example. The car makers today in Japan, in Europe and the United States are practically the same companies. Except that in Europe today, 50% of the car market is made by diesel and practically 0% in the United States and 0% in Japan. The only reason is incentives and regulation. That's the only reason. Consumer looks at price of the car, total cost of ownership, resale value, safety, that's it. At the end of the day, the regulator has a lot of responsibility about okay, I'm gonna drive you more towards electric, power diesel, power gasoline, power flex fuel in Brazil with ethanol. So that's why we need to work together because at the end of the day, consumer shows choice, important driver, regulator also. So given that, where are we likely to see this happen first? Paul. Where are we likely to see which stage happened first? I mean, we're already seeing. Well, the most aggressive adoption of autonomous and probably Silicon Valley because they're used to this beta testing. Because they want to do it. No, I think it's where the question I was talking about will somebody set aside places where it's controlled? And my guess is that that it will happen. In fact, as commissioner was saying, in places where there's already a controlled environment. So it may be that housing developments get built specifically that have set ups for autonomous vehicles in them or a city center gets set aside or a corporate campus gets set aside. I think that's where we're gonna see the full driverless capabilities coming first. I wanted to just go back to one thing about cybersecurity. So we're already addressing this issue in the smartphone and the sort of the state of the art thing to do now is to use machine intelligence that actually sits inside the device and watches the behavior of the device to see does it start doing something strange? Is it going out to some IP address that it shouldn't go to? Are sensors on that shouldn't be on when it's in a certain state? So it doesn't stop the hacker from getting in. I mean, it cries. But it watches to see and make sure it spots it as soon as it happens. There's multiple layers of defense for sure. You want to keep them out, but when they get in because they will, then you want to be able to watch the behavior of the system to see is it doing something strange? Now, of course, there's going to be a cat and mouse game between the attackers and the defenders. So now you're gonna have another AI that's gonna try and hide from the one AI and we'll have a whole game there. And that's a, we could go off on a different panel on that issue. So I'll stop there. If we're gonna mention hacking, we should start talking a little bit more about the dark side of what we're getting into here. And I wanna take this conversation into another issue that's going on, which is automation in general and job replacement. And the question of whether that's gonna all happen smoothly or we're going to see various elements of the society rise up against it. So we've already seen a situation in New York State where the Uber drivers now want to demand that Uber guarantee them five years because they have to buy a new car and they have to amortize it over that period before they get replaced by a driverless car. But consider the situation when we have massive numbers of driverless trucks being introduced. Is that something that the truckers are just going to allow to happen or are we going to start to see protests against that? And you can imagine all kinds of fascinating things that could happen. For example, hijacking four driverless trucks and shutting down a highway. In other words, these are the kinds of scenarios we could get into because driverless cars, driving is a social activity with ramifications for every aspect of society. Well, it brings up an interesting point because obviously we've been through those kinds of adjustments before. I mean, half the population worked in agriculture a hundred years ago and no longer does. But what's interesting about this is just the speed of the change. I mean, Carlos, it seems to me the auto industry until recently pretty much was operating on the same model it was operating on 100 years ago. And all of a sudden, as you just said, now you're going through an electricity revolution, an autonomous revolution, a ride sharing revolution, a connectivity revolution, all at the same time. I mean, the ability of industry and society and regulators to deal with change at that pace is sort of mind boggling. Yeah, but at the same time, it's a huge opportunity because at the same time you are reshuffling the cards and you say, hey, there is a new game, let's see who's going to be able to prevail. So I think a lot of risks, a lot of opportunity, but I can tell you that for car makers and car makers who can invest and betting a lot that the car is not a commodity, but a very intelligent, it's going to become more and more an intelligent product, it's a great opportunity because you're moving away from the commodity. You're moving from a transportation device to a mobile space, which is going to be smart. It's a huge difference. And even the value of the car, even for young generation, I'm always asked a question, ah, young people are not interested in cars, they're not going to be buying cars, we're seeing it in a lot of developed country cities. Okay, but at the same time, because they want connectivity, that's what they want. Okay, I'm not ready to stay in a car for one hour, two hours, do nothing. Now with everything coming on connectivity, where you're going to be able to video conference and text, play games, see your movie with the autonomous drive or a car without the driver, well it becomes a very interesting space. On top of this, for the moment you cannot move your house or your office, but you can move your car. So you can be in the car and do everything you're doing in the house and you're doing in the office, except that it's mobile. That's why you have so many company interested in the car. You're saying, oh, all of a sudden, Google, Apple, all these guys are discovering the car industry. No, because they are seeing the fact that this is going to become a very important object in the life of the company. Big part of your life. Exactly. So now people say, okay, so then the car industry is gone. These are the dinosaurs of the, not at all, why? Because the car, in difference to all the other devices like the telephone, is the most regulated object, the most regulated object. Because at the same time, that has to be connected. It has to be autonomous. It has to be zero emission or whatever. It has also to offer attractiveness. It has to offer comfort, driving pleasure, safety, and all of this. So it's becoming more and more complicated. And we're looking forward to it. Let me get your perspective on that, Paul, because I work in an industry when the media clashed with technology, technology, the technology industry won. Google and Facebook took all the profits and we're now struggling with the remains. The same thing is happening now in automobiles. Where is the value going to be created? So just on the job side, I think that once the car is relatively autonomous, you might actually get, and there's car sharing, there might actually get cars that are for certain services. Like a car might drive up where I get my hair cut or I might get my nails done or I might have a doctor's appointment or something like that. I mean, we might create all sorts of jobs because now there's going to be this very distributed set of services out there. Creates new work opportunities. I think so. I mean, we shouldn't say, things are going to go the way they are right now, and there's going to be some really cool stuff. The question with the speed of change is when somebody comes, when trucks, right, I'm told they're the number one employer of non-college educated males in the US. When those people are no longer driving trucks, they're not necessarily going to be equipped to do mobile nail salons or whatever. I mean, some people will get, look, there will be certain skills that people will be able to be trained to do. I think some of these are higher skilled and some of them probably are a little easier. I wouldn't expect the truck driver necessarily to become my doctor, but that may even happen because we're building all sorts of systems that are going to allow consumer level people to do diagnosis. I mean, we are doing this tricorder X prize right now to do exactly, so we don't know. A lot of stuff is going to change and to take where we are today and project that out in the future and say, okay, there's nothing new coming. I mean, I can even, I can sit here and imagine lots of things that are going to come. But, and then let me just go back to the collision between the car companies and the, or collision, the collusion, whatever it is, between the car companies and the technology companies. Is it harder for a technology company to learn the business of making cars or harder for a car company to learn the sophisticated technologies? My experience when industries converge is that neither side really gets it and they have to spend some time to understand each other. Great example is in healthcare. When we started out with that a long time ago, all the people from the healthcare industry thought they were going to become a mobile virtual network operator because they knew their industry was hard but they thought the other guy's thing was easy. And everybody from the wireless industry thought they were going to be selling diabetes test strips because that's got to be a really easy one. So it took a while for everybody to come together. I think we are doing that right now and I would hate to see that there's this polarized thing where people actually aren't working together. I do think that those connections are being made though. Right now I would like to say just to add to the fact that we have to be very modest, prepare for the future but not try to preempt it because you don't know is the choice of a car is not only a rational choice. It's also an emotional choice. I mean choosing a car is something between choosing a refrigerator and choosing a pet. You don't select a pet on a rational base. You select it for many other things and the car is something in between. So we have to prepare still to create and continue to have this emotional link that you see clearly when particularly you go to India, Brazil, Russia, China, see the ceremony of buying the car and all the families coming and everybody remembers exactly what was the first car in the family and the link to the brand. This emotional side is gonna continue. On top of this we're gonna build all this rationality particularly coming and that's why we have to be prudent not to preempt too much about where the consumer's going but being ready to respond with a full menu. But I would like to jump here here because there is another dimension that we should not overlook which is a political environment in which all this will be happening and the social issues are important. We don't have at this point really well designed tools for reeducation of all this massive change that have not been experienced before because digitalization is really making dramatic shifts and it used to be in one lifetime that these shifts happen but now they're happening two, three times in your job lifetime. So I think it's gonna be also elections that will have a great impact on that because people will be of course lobbying and politicians will respond. So all that will kind of have an impact. I'm saying again, there's no doubt where the development is going but in order to define the speed all this will have to be taken on board. One more thing if I can just quickly say we are forgetting that we are moving also to a completely new concept of mobility. We call it mobility as a service where multimodality will really play an important place and automation has been known to public transport already. I mean there are many autonomous and drive less public transports already available in capital cities. So this is not a completely new experience. It's a completely new personal experience but we've been using it. So mobility as a service will also play an important role and it will be a systemic solution that will come out and the cities will be looking at the system solutions more than just individual solutions. So all that will impact the whole acceptance. Let's take some questions. Go ahead, you first. Right here in the second row, yes. And identify yourself before you ask your question. Gary, going with BD. So first, this has been a fantastic panel, the energy of this discussion of what I think is still a little bit of an unpredictable long-term picture but if I can ask one thing, let's no longer keep saying that in an autonomous car where the driver may be playing a video game or doing their emails or brushing their teeth might be intoxicated, they may feel safe at that point or may doze off that right at the point of an emergency the control goes back to the driver. It's something we have to stop saying cause it undermines the credibility of the whole principle or that it's gonna go to somebody in a robot tower who the NASA analogy, that's one spaceship going to Mars. Here you're gonna have hundreds of millions of vehicles. It may go to voicemail. We have to have an explanation other than that. We're gonna go buy a Model T and only drive on Sunday mornings at five AM because I'll be afraid to be on the road. One other quick point I wanna make. The estimates of job loss in the United States are four to five million associated with trucking, courier services, taxis, the absorptive capacity of any economy to redeploy to other skills and other jobs which is already under great question even based on today. This has to be taken very seriously as one of the questions in terms of the societal trade-offs. There's two elements on that. I think actually the trade-off issue which is one that many of us have been talking about for about a decade has been more or less settled. I think it's understood that the trade-off isn't gonna happen easily. The driver is not gonna be held responsible and we're either gonna have to have some kind of no fault insurance or the car companies are gonna take responsibility for that and that will get factored in by the benefits that they see the cars will bring. So we want to have trade-offs particularly when we're talking about autonomy per se but that's different from saying that the drivers are gonna be held responsible if something goes wrong. So I think actually we're beyond that particular issue. The question of tech unemployment and I mean that's just one example of where this is converging with so many other issues. When you talk about regulation it's not just driverless cars. We're dealing with a plethora of technologies that really- Well every industry has a version of this. That really our existing governance systems are not set up to deal with and we need some new more flexible form, new more adaptive form of governance but in relationship to the tech unemployment question there's a lot of outstanding questions about how many secondary jobs will be created by these new technologies. They're questions about how quickly this is coming so it's one thing to point out how many taxi drivers and truckers there are. It's another thing to wonder well how many years or decades will it take for this turnover to occur. Question in the back. Thank you. Zheng Xin from Caixin Media. This is a follow-up question about industry convergence. So right now the players not only include the traditional car makers Google's and Apple's big tech firms. There are also many startups they are building smart electric cars from ground zero, from scratches. And they are saying we don't have a very profitable old model there that hinders us to move forward. We don't have the old factories and we don't have the reputation like Google and Apple already had. So we can do everything better and faster. You need two to three years to test. We need one and a half. And a more specific question for Mr. Gosu I don't know if you hear about a startup company called Faraday Future in California and they are, you know, I don't know how much we should believe them including FF and others. Yeah, well look there is a big boulevard. I mean particularly you're talking about electric cars. There is a big boulevard for electric cars. Particularly I would say in China. Why in China? Because the Chinese government as you said has stated very clear objectives about how many electric cars they want on the ground in China. And I understand that by 2018 or 2019 they want 8% of the sales of cars be electric while today it's less than 1%. So there is a huge boulevard and where these cars are coming from. Obviously the classical car makers will not be able to fill the 8%. So there is an appeal to a lot of startups. A lot of people saying okay, we're gonna bring our own contribution to that. Which is fair. I mean at the beginning you're gonna have plenty of company, you're gonna have plenty of models, but little by little selection is gonna come from the choice of the consumer. And this is where when price start to come in, services start to get in. This is where the car makers have a bigger advantage because they have a much bigger scale and particularly they have a much bigger network. I'm not saying that it's already written on the wall but I think when the bigger player who have already a network in the country start to go into this direction, they start with they have some advantage. Yeah, exactly, some advantage. I've been in the Faraday Future car. It's a real car and one of the things that it does autonomous or driverless even is it parks itself in a parking lot and that's pretty cool. Just wanted to go back to the driver distraction thing because you listed a whole bunch of things the person might be doing. They might actually be in virtual reality and then their whole context of their brain is somewhere else and they won't be able to context switch back at all. So that goes back to what I was saying earlier. The car has to have very good situational awareness and try and predict when it is that it has to give the driver enough time that they can and of course it won't work all the time so there will be accidents from that. It's a human machine interface it's gonna be the big issue I think. Go ahead. Hi, my name is Eyal, I'm a global shaper from the Tel Aviv Hub and many of the benefits of autonomous cars come from an underlining assumption that will break the model that exists today of one person in one car. This is what causes congestion many of the other difficulties with cars. But so far history has shown every time we have technological advancement in mobility technology it just raises the demand for owning a private vehicle and as we said this is an emotional experience. People would like to still own their own car and brush their teeth there and work there and not have anybody interrupt them. Data today shows that people are not selecting the shared ability option. You showed that in your talk before. Not in Uber, not in any of the other carpool apps that were out there. What makes you believe the assumption that people will choose shared mobility and not just buy more cars for private use and increase congestion and things like that? Carlos you're not gonna be upset if they end up buying more cars. No, no, no, no, let me tell you. No, I'm gonna answer it because it's obviously something we work because it's our business. I mean this 2016 record year for the car industry 91 million cars have been sold in 2016. This includes the fact that in many cities people don't buy car anymore, young people don't buy car. We know this in California, we see it in New York, we see it in Tokyo, et cetera. This includes, which means there is room for the classical car industry to continue to develop particularly based on growth in emerging market. That's what's happening today. And at the same time you have an explosion of demand on shared cars, shared services, et cetera. I don't think it's either or. It's both. You're gonna have both of them and I agree that there is a lot of rationality into choosing transportation but there is also a lot of emotion into choosing cars. And you're not gonna ask people to say, okay, you're gonna have, do one or the other. They're gonna do both. They're gonna say, okay, this is my car, I'm gonna use 5% of the time but I like it. I have a four by four even though I'm driving in a city, et cetera. And then from the other side I'm gonna use more Uber into certain transportation. There is room for both and we need a car maker to address both. Well here the regulators will come in, right? Because the cities are way to congested, we need to really care about the pollution and things like that. So we already have now a rule that the user pay, polluter pay principle, it will be the charging mechanisms for the use of infrastructure will be reshaped. So everything at the end will have to balance. I mean the cities will certainly push for more, for less cars in the cities, they will push more for public transportation and sometimes it just won't be any other choice. But the number of people on this planet earth is growing so we need to take that into account as well and middle class is strengthening in the biggest countries like China and India and of course all that will have an impact. That's why decarbonization will play a stronger and stronger role and will definitely impact also new models. If you use the lodging as an example, I mean apartments did not kill homes that people own, hotel rooms didn't kill the, you know, and we already see some things where there are, it's not an either or there's some people will do this, some people will do that and some people will do both and that. Overall increase demand for mobility, I would assume without any doubt. Right here. Thank you very much. My name is Kumagai and I'm a chief economist at the Diver Institute of Research in Japan. And I think driverless car will change the nature of the automobile industry because now the industry is to produce cars. But in the future, I think the industry might be the transportation industry to transport somebody or something to other places. In that case, considering the urbanization, I think the automobile production will decrease dramatically because the utilization rate of cars is less than 10% now. And I'd like to ask one question to Mr. Gon. Considering such kind of structural changes, will your company survive, sustain the strengths and what will be the core value of your company, of Nissan in the future and how will you beat Microsoft, Apple, Uber, future competitors? This is an investor's phone call now. We're not gonna beat Microsoft or anybody, we're gonna join forces with them and that's what we're doing, by the way. We have an agreement with Microsoft on connectivity and working together on the car. I don't think, I mean, obviously the media likes to say, okay, this is a battle between X and Y and Z. No, this is more a collaborative effort where everybody is trying to bring a contribution. We are working with Google, we're working with Microsoft, we're working with Uber, we're working with Uber and all the other car manufacturers are doing the same thing because we're not looking for the same contribution. Everybody has a different contribution. Now, I don't think you should think about either or. What you're saying is right, that mean urban mobility is gonna change a lot and the way I see it is certainly totally electric, lower number of cars, more shared services, et cetera. But let's not forget that because we are focused on Tokyo, on New York, on Paris, or on Berlin or London, we understand that. But let's not forget that in China where you have more than one billion people, in India where you have more than one billion people, in Africa where you have also more than one billion people, you have still very low level of motorization. In India, you have 20 cars per 1,000 inhabitants while in the United States you have 700 cars per 1,000 inhabitants. We cannot say, oh, I mean, we're gonna look at India the way we look at the United States, it doesn't work. Even in China with the boom of cars sales in China, you still have 150 cars per 1,000 inhabitants and 70% of the car buyers in China are new buyers. They never own the car. You're not gonna stop that. Africa, we are at the beginning. You have 10 cars per 1,000 inhabitants in Africa. So you're gonna have this motorization which is gonna continue from one side. And I don't think anybody's gonna stop it because this is a signal of autonomy. This is a way to get education. This is a way to get a job, et cetera. But at the same time, we need to face another situation which is mainly an urban situation in developed country where you have more services. So car makers are not gonna become only mobility provider. They're gonna be car makers and mobility providers. I don't know, have you considered also the fact that there's going to be more demand for mobility because there are classes of people that don't really get out of their house like senior citizens and in developed countries where we're seeing demographic shift in much older populations? I actually think that might be a source of demand. I don't know whether you've modeled that at all, but for sure, senior citizens are going to get out of their houses in cars that take care of them and drive for them much more, I think, than they do today. Or are those people more willing to go places if they can do something else while they're in there? Yes, they can, yeah. Go ahead. I'll read the panel from South Africa, Mubanka. To what extent will the rollout in developing countries be very much slower than developed countries because I come from a developing country and our infrastructure is often inadequate, road signs are not replaced, road markings are faded or non-existent, traffic lights frequently not in operation. And how does that affect the speed of rollout of this new technology into developing countries? No, I think the answer is yes. We are, all the tests that are taking place in Tokyo, in Palo Alto, in Boston, in Paris, why? Because we need, in order to test this technology in infrastructure, and we need reliability of the existing infrastructure, a new infrastructure, and particularly a very good adequation to the laws of driving. So that's going to give a lot of advantage to developed countries. Now, this being said, when this is going to be tested and mass marketed, obviously it's going to develop in emerging market, but we're going to need a lot of infrastructure. And that's why I don't think we can do it alone. I think we're going to need to discuss with the local authorities about how much they want this technology in their country or in their cities. And obviously they're going to have to bring their fair share to this. And remember, there's a lot of connectivity even in developing countries. And for places where there isn't connectivity, we're actually in the process of building a low-earth orbiting satellite system that's going to give you broadband everywhere, including all of those places. So there is going to be the ability for sharing of information, so crowdsourcing of mapping data, and things like that as well. So there will be some technological solutions even in places where it isn't existing infrastructure. I wouldn't underestimate what the colleagues pointed out, because the capacity of investments will play in a really important role also of the deployment of these technologies. And probably the countries that are still trying to bridge the development gap will hopefully learn from other countries and maybe invest smarter. And really see the big picture and use the systemic approach at the beginning and not trying to repeat all the steps that were done in the developed countries. My name is David Bahou from the French newspaper Lysico. A question on affordability of those cars. This future looks really good, but cars prices are more and more expensive every year, even though they are low-priced cars. Like phones. How will people be able to pay for this? Do you expect the economy of scale to be very rapid and that the prices will go down? Or actually, the car will exist, but nobody will be able to really pay for it? No, the affordability is at the center stage anyway of our companies. Well, you know, for example, if we take, I'm going to take an example today, which is electric cars, OK? So the attention are more on the sophisticated premium electric cars, which usually have a sticker price of $60,000 plus, which represent a very small portion of the market. We are addressing the core market. We're talking about driving the price down in order to really address the mass market. So and we see a lot of response. So what I'm telling you, there is more visibility for the expensive solution, but there are a lot of solutions which are affordable. And affordability for us is the key to the mass marketing. Let me give you an example. In China, what is selling to the, China is the market which has grown the most in electric car in 2016, more than 100% increase. And what is growing are the low cost electric cars. It's not the $80,000 sticker electric car, which is obviously make front page everywhere. But this doesn't represent anything in terms of how much contribution you're bringing to the lower emission. You need to address the low cost, the mass market, and this means affordability is center stage for us. Yeah, we're at we're going to say car sharing can help in that as well because it will amortize the cost among many more users and don't underestimate how much your smartphone is going to do and be able to interface with a car at the very low end. We're out of time, but thank you. Thank all of you. Very enlightening panel. Thank you. That's great. Thank you very much. Thank you very much. First a pleasure.