Automated vehicle innovation and update by consumers. The motivation here is pretty simple. If these vehicles are in fact safer, and i think we all think they will be, that any policy that results in unnecessary costs or delay will result in additional property damage, injury and death. So death by regulation is really something to be concerned about. And i know for a fact that there are people within the federal policy making world, are cognizant of this fact. So what i think nhtsa should do is focus on the federal Motor VehicleSafety Standards that may innovate techno lodge innovation in our automated world. An example was this petition from tesla motors to nhtsa regarding fmbss 11, and they were seeking to comply with the mirror rules with cameras, rather than using mirrors. Even if tesla were to replace all mirrors viewing functions with cameras, they would still be required to install mirrors. I think this is an early example. They just want to have the option to comply with mirrors. I think well run into a lot more of these Going Forward as Automated Vehicles get more and more advanced. So if congress is to do anything, and i dont think they should do anything at this point, perhaps not even in the future, they should have nhtsa report on these potential conflicts. Im not sure what well get out of this. At least its a starting point and its something that congress can sort of use as a jumoff point. Now, like i said, theres already some examples. And this quote is from d. C. Council member shay, her criticism of my bill that ultimately had passed. Ill get to why this is flawed in a second. She was under the belief that it was the technology that was currently then, that may have been true briefly. Its certainly not true now. I think youll see why in a second. So im going to use washington, d. C. , were all here, as the case study. 2012, this bill is introduced. The original bill from Council Member shane, all vehicles be powered by alternative fuels. Impose a new mileage based tax on all Autonomous Vehicles. If you can see, theres a logic here. Council member shay got a ride in one of googles Toyota Priuses and realized, if all of these Automated Vehicles are going to be powered by alternative fuels, theyre not going to be paying the fuel tax, so we need to tax them somehow. Theres a logic there, i just dont think its a good logic. And mandating a licensed driver be in the seat with autonomous operation. That driver seat rule remained. If you look at the april 2014 proposed rules, from these, youll notice that they seem to require that an operator have a special d. C. Drivers license endorsement. So the implication there is that the district is calling for or calling that all test drivers be d. C. Residents and have a drivers license issued by the district of columbia. Given that we live in a metropolitan area where most people live outside of the district of columbia, even from a regional perspective, that seems to be restricting your potential test driver pool unnecessarily. So i think thats a sort of a ridiculous requirement. And hopefully the final rules wont reflect that. But, you know, beating up on d. C. Is easy. California and michigan also have these driver seat requirements. So california, ill focus on them now. So, california, theyre another place they rolled out their manufactured testing rules earlier this year. They imposed this drive seat rule. As randal showed you that video of the latest prototypes, google developed this pod car, lowspeed vehicle. And they want to take out the Steering Wheel and pedals and basically be fully automated. The operator will have no ability to retake, or take Manual Control at any point. But state testing, and federal lowspeed vehicle rules require the installation of the Steering Wheel, all this stuff they dont want. What we have here is a regulator promulgating a rule forcing the innovator to take a step back. Were already having negative impacts from regulation of Autonomous Vehicles. This is really unfortunate. I dont think it was intentional. It just goes to show, you know which roads good intentions can pave. So the vehicletovehicle Communications Mandate that randall mentioned, nhtsa did issue an advanced notice of rule making in august. What they wanted to do is they wanted to develop a final rule by the end of the decade that will require all new highway vehicles be enabled with dedicated shortrange communications, vehicletovehicle technology. And this mandate wouldnt be requiring that all cars have these little yellow circles around them. What the d. O. T. Image is trying to illustrate is the cars will be talking to each other. They wouldnt be completely reliant on onboard sensors. For instance, in finding hazards. And keep in mind, this is not for automation. Theyre talking solely about warning drivers of hazards. Imagine theres a car, a few car lengths ahead of you, slams on its brakes. This would send some data back to your car and tell you, hey, some guy slammed on his brakes ahead of you. Take whatever action the car is going to tell you to take. It remains to be seen how effective that will be, or how drivers would actually respond to that in real world settings. So theres some problems with the vtov mandate. The proponents argue its low cost. Theyre imagining that all new vehicles will have these vtov boxes in the car that will communicate with each other. Others say it will cost a few hundred dollars. Whats the big deal. Were looking at 25,000. So this isnt a big increase in the sticker price. The problem is, that the benefits are likely to be very low in the short run. Were going to have these things are at earliest, were expecting a mandate by 2020. Its going to take at least ten years before you have significant Market Penetration for the auto fleet to turn over, that this is going to do any good for anyone. I mean, i think you need about 70 of vehicles enabled for you encountering one on the road to be greater than shant. We expect it to take at least ten years, if we proceed as nhtsa appears to want to do right now. Dsrc has had this 75 megahertz has been blocked off since 1999. The proponents obviously want to keep it blocked off for transportation purposes only. But there are forces out there, namely those who make consumer electronic or portable Electronic Devices that are wifi enabled that would love to have access to the spectrum. Theres a battle going on at the ftc that hasnt been resolved yet over the spectrum. I think that reason alone, i think it makes premature for nhtsa to dive into this, but they thought otherwise. They ignore competing technologies. Nokia earlier this year announced it has developed an advanced form of lte, that would be able to perform these same vehicletovehicle communications functions. But you would be able to use technology you would be able to use things like cell phones much more easily. And we already have an lte Network Infrastructure out there. When we get into sort of more precise vehicletoinfrastructure kind of applications, that would require, if were dealing with dsrc installing a lot of roadside boxes, and we just frankly dont have the money for that. Nhtsa has ignored some of these technologies. There is a chance its already obsolete, too. If were talking 2030, 2035, that, you know, what good is at least the way nhtsa is imagining this, and not that the context of automation is not bad. If you have a fully a selfdriving vehicle, we have no responsibility to take control of that vehicle at any point. Perhaps not even the ability, what good is a warning light, or audible alert going to do for you. The answer is not much. Theres some unanswered cybersecurity and liability questions. This proceeding just opened an industry thats very interested in these. Theyre very concerned. As far as an automation specific cybersecurity potential problem, well, if you have automation based on sensors and onboard computers, how would a vtovk mandated vtov system, its onboard the car, how will the two systems interact, if at all. So thats why i think the best case for fully Automated Vehicles under this vtov mandate as currently conceived is automakers will be forced to install completely Useless Technology in Autonomous Vehicles. That might not be bad, but that would certainly increase the price. And that gets back to unnecessarily delaying the rollout to the consumers. I think the takeaway for vtov and dsrc, how nhtsa currently envisions it, we should be skeptical. Vtov is so valuable automakers would not consider installing it voluntarily. There may be legitimate institutional problems with industry collaborating on standards. But then perhaps those should be addressed directly, rather than forcing a mandate down everyones throat. So im going to end with some sound Automated VehiclePublic Policy principles. That i think are sound principles. I think we should start with recognizing and promoting the huge potential benefits. I talked a lot about the safety. Randal did mention these traditionally, mobility disadvantaged populations, the disabled, the elderly and the youth who dont have access to the mobility many of us take for granted. We should reject the precautionary principle. This is a new technology. We dont have data. It doesnt mean we should shift the burden to the innovators and say, you need to prove that this is safe. Before we get these to consumers. I dont think any auto manufacturer is going to release to the Consumer Market until theres some demonstrated safety level. But i think going down a precautionary principle route is a good way to keep these out of the hands of normal people. We shouldnt presume to know how the technology and law will evolve. The technology right now, were talking about proprietary things. The only thing we know about this is generally the kind of the press releases we get from industry thats currently developing this. And the law theres no court cases yet. We dont know how this is going to do we need legislative intervention to update our liability laws, things like that . We dont know. Theres certainly a possibility that common law liability can evolve without any sort of intervention. And then, i think, you know, number four logically follows, we should always seek to minimize legislative and regulatory intervention. Regulators are slow, however well intentioned. This is moving pretty quickly. We should let the innovators innovate, and keep the regulators as far away from these things as possible. Doing maybe some very minor things around the edges, but they shouldnt be involved in the sort of developing the technology. Or deploying the technology. And then finally, once we get to the state where were talking about, you know, updating Motor Vehicle codes to reflect this new automated world, and really doing some serious legislating, if it comes to that, we should focus on developing clear and simple rules that have neutrality. I see a big risk of the first mover. So the First Company that ends up with a consumerready vehicle coming to market, and then regulators saying, well, were done, this is the technology were going to mandate, i think its a terrible idea to mandate technology in generation one, despite the fact that regulators may believe that this will enable to get generation one tech out to consumers more rapidly. So with that, i am finished. And i look forward to any questions. Thank you very much. Thank you, marc. Our next speaker is adam thierer. Hes a Senior Research fellow with the policy program at george mason university. He specializes in technology, media, internet and free speech policies. His writings have appeared in the wall street journal, the economist, washington post. Adam has authored or edited eight books ranging from media regulation and child safety issues and High Technology market. His latest book is the continuing case for comprehensive technological freedom. Adam was president of the progress and freedom foundation, director of telecommunications here at cato. He got his masters in Business Management at the university of maryland. And his bachelors in journalism at indiana university. Adam . Thanks, matthew. And great to be back here at cato. Its always a pleasure. Its also a pleasure to follow randal and marc. I learned a lot from what theyve both written on this, and many other issues. And i recommend it to all of you. My remarks will be focused primarily on the privacy and Security Technologies in Driverless Cars. You can download my paper from the cato website. In thinking about privacy and security in smart cars, or Driverless Cars, we need to begin first by understanding, acknowledging that security and privacy are relative concepts, with very amore fosed boundaries. Not everyone affixes the same values on this. Some people are hyper cautious, and hypersensitive about the privacy. Others are risk takers that are just somewhat indifferent or more pragmatic about their privacy. We all say we love our privacy, but sometimes we do things that actually in the real world act differently. This is called the socalled privacy paradox. We should understand the security and privacy norms tend to evolve over time. And do so very rapidly. With any new highly disruptive technology, such as Intelligent Vehicle Technology, we often panic at first. Especially about the privacy situations of the new technologies such as these. But we establish new ethical and legal base lines about new technologies fairly rapidly. Ive written about this in recent law review articles and in my new book. Theres essentially as i describe it in that work, a cycle of work of initial resistance to a technology, gradual adaptation, and then eventual assimilation of that technology into our lives. Not without some heartburn along the way, of course. This was just as true for the first cars that came along over a century ago and it will be true for the new smart car technologies that are evolving today. But it is especially true that these norms have all as it pertains to privacy and security, that theyre subjective and relevant. Third point, for almost every perceived privacy or security concern or harm, there is a corresponding consumer benefit that sometimes balances those out, or even outweighs those perceived harms or fears. We see this reality at work with the broader internet. And with digital technologies. And well see it at work with shall driverless vehicles. Compare todays telemattics and Intelligent Technologies one board our cars with the tracking technologies onboard every smartphone that all of you are carrying with you right now. The reality is, these technologies, the ones we carry in our pockets, the ones well have in our cars and already have in our cars today, are capable of tracking us. That sounds sinister. Tracking, of course, creates enormous benefits for us as well. We now know in realtime what traffic looks like when were in our cars, not just because the technology in our cars, because, of course, we can see it on a map. In realtime that happens because were all connected and were all being tracked. So again, we have to be clear that not every theoretical boogieman is just a big bad awful thing to be disposed of. It has a corresponding benefit we have to take into account before we address it. Fourth point. As it pertains to intelligent vehicle technologies, todays privacy and security concerns are not the same as yesterdays and theyre not going to be the same as tomorrow. Todays intelligent vehicle technologies are likely to be more pressing i would argue than tomorrows. Thats things like event data recorders and telemattics are recording our realtime actions today as we are doing them with our hands on the wheel. This raises a variety of interesting questions, which is, can that information be used for automobile insurance purposes. And it is today already on a voluntary purposes. Or breaking the law by driving erratically or too fast. They raise obvious privacy considerations. That may even lead to considerations or concerns about Discrimination Associated with the data thats collected by our vehicle. But its worth asking this question. What happens as we make this transition to fully Autonomous Vehicles. And what happens when our cars are less of a final good that we own and more like a service that we just use or rent on occasion. What happens when we combine the power of todays sharing economy with the power of selfdriving technology. A car of the future may look like a robotic chauffeur, as randal described. Theyre just waiting out for us, we dial them up on our cell phones as we need them. Needless to say, in that environment the privacy considerations are very different than they are today. Clearly we can still be tracked, our activities can be tracked, but not us personally manually operating that vehicle, and so its a different type of consideration. Thats something to take into account. Fifth point. This is the most important point ill make here today, that any security or privacy solution should take into account the source of considerations that ive just outlined, and must be able to accommodate the many different types of views and values that people have as it pertains to privacy and security. I write about this in my recent book, there are no Silver Bullet solutions to concerns about safety, security and privacy. It will be very difficult for law to keep pace with not just rapid innovation in this space, but the rapid evolution of consumers taste and values. This is why we need a flexible layered approach to addressing security and privacy concerns. We need to borrow a phrase from Richard Epstein of chicago, simple rules for a complex world. We need to rely on things like contracts, and enforce existing terms of service. We need to think about how common law and torts might evolve, or specifically products liability. For that, and i want to highly recommend a wonderful new paper from the brookings institution, where he talks about the law historically and say its not expected to stop there. When confronted with new often complex questions involving products liability, courts have generally gotten things right. Products liability has been highly adaptive to new technologies in recent decades and it will be emerging as the need arises for Autonomous Vehicles. I agree with that. It will be interesting to see how the liability and Insurance Standards evolve. Again, as we move from our vehicles being final goods to services. Because this will change what economists call the least common im sorry, the least cost avoider in these situations. Right now, you as the driver of your vehicle, are the person who can take action to avoid potential types of harms, whether they be safety related or otherwise. Maybe privacy and security related. But what happens when you no longer own your vehicle. What happens when your vehicle and the people who created it know more information about the vehicle than you do. The knowledge and therefore the responsibility and liability moves from you to them. They already know this. Theyre already thinking about this. This is the least cost avoider principle in economics. We shouldnt be surprised to see legal norms come to reflect that over time if we allow common law to evolve spontaneously. Now, in the process, there are steps that manufacturers of Intelligent Vehicle Technology could alleviate a lot of headaches for themselves. Especially as they come to deal with more liability eventually. And this leads to a lot of proposals for socalled privacy by design, or security by design. The idea of baking in best practices regarding data security, or Data Collection, or data availability, in systems as you manufacture them. This could include things like Data Collection, minimization, or limitations, basically not allowing third parties to access certain types of data that are being collected about you and how you operate your vehicle today, or about how its used in the future. It could include best practices regarding better transparency about how your data is being used, if shared with other parties. Potentially, clear use practices about how you use these things appropriately. With better information or empowerment or education of the users. And clear consent for any types of new uses or et ratiiteration technologies as theyre evolving and more innovation is occurring. You might want to take a look at what the Privacy Forum is coming up with in this regard. This is an ongoing process. There is no end point in this process. These standards and these best practices evolve not just every couple of years, but every couple of months. Because theres always some sort of new technology collecting more data, that raises still yet another privacy or security concern. Heres a query here we might have. We might say, this is all fine and well, but shouldnt there be a minimal Legal Standard . Shouldnt there be a state or federal regulation regarding driver privacy or security . After all, these values are pretty important to a lot of people. But i think for all the reasons ive already stated, we now know especially in light of what randal and marc have already presented, that these technologies are a fastmoving target. I mean, this is very, very hard for me to imagine that any law we put in place today will be applicable to whatever is down even 18 months down the road. You need to have that flexible ongoing evolutionary approach of voluntary best practices, along with the evolution of common law, to probably better address the evolving cases and controversies that will develop in this field. As marc and randal already noted, were probably better served by having a waitandsee strategy, precautionary principle based approach, and instead we should continue to have our default be innovation. This is the point i do the plug for the book. Continuing case for i argue the benefit of the doubt should be given to those innovating, and innovation more generally. Ongoing experimentation and trial and error will permit us to find new and better ways of doing things. New and better ways of being safer, more secure, and even potentially more privacy. We cant be living in fear of shortterm worst case scenario, or else longterm best Case Scenarios will never come about. So one final point i want to make about privacy and security. Especially privacy. Special consideration needs to be paid to the role that government plays in this regard. Government actions can affect user privacy obviously in a very profound way. Many of the privacy and security concerns regarding private sector Data Collection with regards to Intelligent Vehicle Technology are problematic, and obviously deserve some discussion, governmental Data Collection and use is an entirely different level of concern and set of issues. They cannot tax you or imprison you, because they lack the powers the government has, although it is possible to ignore or refuse to be a part of certain types of technologies, including driverless technologies, you dont have to use them, the same is not true of governments who can obviously be abated. Special protections are needed for Law Enforcement agencies and officials as it pertains to the technology. When government seeks access to Data Collected from these technologies, strong constitutional and statutory protections should apply. We need stronger Fourth Amendment constraints. And the courts are going to need to consider revis tinge the thirdparty doctrine which holds an individual sacrifices his or her Fourth Amendment interest when they divulge it to a third party or share it for whatever reason even if that party promised to safeguard that data. What we want is a world where many different intelligent Vehicle Companies are competing on security and privacy, and maybe some are offering those who are more privacy sensitive better protection than industry norms or best practices. Industry is already moving in this direction. We had a recent jao report that surveyed ten makers of these technologies and found they are all taking steps to protect these things to some degrees. But more could be done, and more is being done especially with the liability question looming large. When the government comes in and says we need back doors into all these systems, were having this debate right now with our smartphones, right . Apple and google start using better encryption on our smartphones. The government is saying, no, you cant do that. A lot of people are clamoring for better security on their smartpho smartphone. Thats the exact same debate that will unfold with the smart car and driverless vehicles. We need to allow innovation to go forward. Thank you very much. Thank you, adam. And thanks to all of our speakers. We are now going to enter the q a part of the event. A few notes before we begin. Wait for the microphone. We have a few interns here to help out. And before your question, please announce your name and your affiliati affiliation. And can you please make your question, a question. This is the questionandanswer part of this statement. If you would like it directed at one of the speakers, please make it clear who you would like to ask. Ill begin with the gentleman at the back. Is the microphone on . Thank you. Mark carr. I have my own company which works in freight. Im actually here thinking about a role that i play, transportation research. And for the last speaker, you talked about privacy and all. Im involved in freight. And im wondering where in freight, and whether it comes to our homes, or whether thats commodity freight going back and forth between ports, where do trade secrets fall in with the privacy . Because you can envision shippers who dont want other folks to know what their trade patterns are or their customers and such are. Thats a wonderful question. I have not spent a lot of time researching it. I think ill have to now. Im always careful to wade into the waters of intellectual property, patent and trade secrets, its a complicated field. I would hope through ongoing experimentation with different business models, that Companies Find ways to protect those trade secrets. Although it may be difficult. Having fleets of freight vehicles that are robotically operated raises a whole host of security questions about who has access to that data and information. Thats part of the spectrum upon which companies are going to be competing, to offer greater security to make sure those trade secrets dont get out. It deserves more consideration than im giving it here. This gentleman at the front, please . Im tom curry, a reporter with Congressional Quarterly role call. I talked to the head of the contra costa, california, Transportation Authority last week. They have the new test track for the mercedes Autonomous Vehicle in their county. He said one of the things in the future benefits that he sees is that these vehicles could benefit to the county, is that these vehicles could go out and detect potholes, and immediately transmit the data back to the contra costa transit, or any trngs authority. The fleet of vehicles would be communicating information in realtime, so that they could maintain the Transportation System better. I just wondered for mr. Scribner and mr. Otoole, whats your view of that, and the cost also of transmitting all this data back to a county Transportation Authority . I dont see anything wrong, if he wants to get that information, for him to design an app, and upload it to apple and android, and people can volunteer to download it. And if theyre driving on a road thats unusually bumpy, the app will detect the vibration and transmit the information to the Road Authority that theres a serious problem with the road. Theres no reason why it has to be mandated. Or why it has to be applicable only to driverless vehicles. Yeah, i agree with that. But sort of going from there, i think the it is really valuable, particularly in the context of Automated Vehicles to have our roads in very good condition, particularly if theyre solely sensor based. These arent things that can vehicles presumably, at least initially, wont be communicating a lot of data between each other, and certainly not with infrastructure. So that would be great. I dont think a mandate would be needed. But certainly, yeah, thats i think thats a very good use of that technology. The gentleman who had his hand up earlier. Thank you. Lawrence gasman, small tech publishing. Im interested in the time frames of this. Randal otoole put up a slide that showed the different types of intelligence the car could have. I think that came from some government agency. But im particularly interested in the advanced ones, because you have cars that can park themselves now. Its fascinating, but it doesnt seem like a huge regulatory issue. But if you put your dogs in the car and they take themselves to the vet, thats a completely different deal. It will be quite a few years before you start seeing that in any, you know, outside california, i guess, is what im thinking. What sort of time frame do you think those will be . And i guess the question is, how long do we have to get this right . Well, it depends on who you ask. Certain automakers are saying that toyota is saying, were not going to focus on this really at all. But then you have Companies Google is still saying 2017, presumably level three, level four. Thats still a goal theyre sticking to. Nissan is saying 2020. Presumably level three, possibly level four. Continental is also saying 2025. So it could be sooner rather than later. And, you know, it really does you know, we need we dont necessarily need special regulations for this, but it just goes to show that regulators are going to be well behind the technology developing. Assuming these optimistic industry forecasts are true. About 200 experts in the field were surveyed on this question in a conference last summer, and the median answer was that google Driverless Cars, a car that has a Steering Wheel, but the car can take over pretty much completely in all conditions, will be available by 2020. The pod car that doesnt have a Steering Wheel, or any way that the driver can control it, other than start and stop, will be available by 2030. I would go further and say by 2040, well Start Talking about closing roads to humandriven cars, because theyre too dangerous. Quickly, i think theres an important distinction to be made between highway and nonhighway vehicles. Certain things that are happening, namely city mobile 2 in the eu where theyre focusing on the lowspeed geographically restricted almost paratransit kind of vehicles. Those you could perhaps see earlier. They would be deployed in college campuses, retirement communities, places like that, where you dont need to meet these stringent, say nhtsa guidelines. So theres the potential for that, too. Happening before we see the actual highway vehicles. This gentleman in the front. Thank you. I was wondering about insurance implications, especially when you have automated and manual vehicles. Has anybody looked into that . Are there any insights . The interesting thing is Insurance Companies are some of the biggest backers of selfdriving cars, because they figure its going to significantly reduce problems, and accidents. One concept ive heard is that instead of when you buy a driverless car, instead of buying an insurance policy for the car, the auto company will buy the insurance policy, and that will just be included in the price of your car. So if theres any liability involved, then the auto company wont have has its own insurance to deal with, and it wont have to deal with extended lawsuits and so forth. Way at the back, the woman in the back, please. Sharon, voice of moderate. The russian monitoring stations that they wanted to build for gps, that was luckily quashed, i believe last december. Now, if people in foreign governments have access to the navigation, and then you have the Driverless Cars and Companies Like nissan own auto vans which is putins car company, theoretically they could drive one of our Nuclear Scientists off a cliff. We start thinking about this technology and security. We do not now have the monitoring stations, which would be the russians having that access. But this tech, i think with foreign companies, any concerns or comments . The concern about Intelligent Vehicle Technology being hacked is very much evident. There have been demonstrations in realtime about how attacks can be perpetrated. Its not something we should take lightly. Its not something, however, we should live in complete fear and dread of. We need to find solutions to these problems. We need to figure out how to make these systems as secure as possible. So that that sort of hacking, for whatever reason, is avoided whenever possible. I will say this, that we have to do this relative to the historical base were operating on, a world where over 32,000 people are already dying behind the wheel because of human error. Thats almost 100 people a day. Will there be hacking that potentially results in someone resulting in an accident or even a death . Potentially, yes. Im not here to say no. But we have to understand that in that world of more Intelligent Vehicle Technology, it may be that we have tens of thousands of peoples lives being saved with that potential of it being hacked at the same time, were going to have to try to make that balance. I know that sounds crude or overly utilitarian, but i think we have to take it to that historical base line. One reason i oppose a vehicletovehicle communication, is that a mandatory system is going to be a lot easier to hack. And its going to be a lot harder to defend against, because the government that mandated the system will probably not be as motivated to defend. Whereas if you have a competitive system out there with different companies, each offering their own Software Packages and other people offering apps and so on and so forth, theyre all going to have a competitive reason to keep their systems from being hacked. And its going to be harder for a hacker to attack, to have a widespread attack. You can attack one particular app, or one particular software system, but you wont be able to attack everybodys at the same time. This gentleman on the left. Good afternoon. Todd wiggins free lance journalist in d. C. I often think of the movie popular some time ago irobot. I think there are some technologies that you have an interest in, that were sort of promoted in this movie. And so i was wondering, are we are the americans most likely to debut this type of driverless car, or is it more likely that South Koreans would be likely to debut . Because often they seem to be ahead of us when it comes to mobile phone technology, the way theyre using items, that we arent really necessarily using just yet. They may be the test ground for those, for technology. Are they ahead of us on this . I think the United States is actually uniquely positioned to be one of the first places to have these available well, in widespread testing, which were already seeing early testing on public roads. But also availability. One reason is were not party the United States is not party to the vienna road convention which has some definitions that may restrict certain testing and operations in other countries. Most countries are party to that. So thats actually really interesting. Theyre actually working on updating the vienna convention, to basically allow for more availability in the future. But it has proven to be a problem. I know a bunch of the this is being led by a number of german parts and auto manufacturers right now. And theyre very well aware of this. As it stands right now i think the u. S. Is in a pretty good position. There will be extra points, if you have a minority point in the next question. The gentleman in the middle. Red shirt. Him. Well, its a great movie, so theres that. Marc, good to see you. I work for techs now. Colleagues ride motorcycles to work, and i wonder how motorcycles and people who enjoy them would fit into a driverless car future . I thought about including a motorcycle video in my show, because andrew of google has developed the driverless motorcycle. Its a little bit like a segway. They can fit in just fine. That was actually andrews first entry, when the darva was still running the grand challenges. It was an autonomous motorcycle. He tells the story, there was a new yorker article about how theyre ready to go. He messes something up at the last minute and it falls over. You know, that was this sort of you can look at the genesis of the google selfdriving car was the selfdriving motorcycle. Gentleman on the right here, please . David sullivan in washington, d. C. I have two questions. This is a very appropriate day to have this panel, because this is the birthday of elwood haines, one of the early pioneers of the automobile whose first gaspowered driving automobile is in the smithsonian. Elwood haines also was the first to get involved in an automobile accident. So i would ask the panel to talk a little bit more about how they see the common law personal injury evolving with selfdriving cars. And my second question is, if in fact people who have time while theyre commuting to read and to do work, and to watch movies, are going to be interested in commuting longer, why hasnt that already happened . Because people already have that opportunity. So those are my two questions. The evolution of personal injury law, and why we havent seen an increase in commuting time already. Ill Say Something briefly on the first one. I already mentioned the evolution of products liability, and personal injury could go alongside that. I think it will be very interesting cases and controversies involving hybrid situations where someone in this sort of vehicle that randal described that is fully autonomous and may be insured by the Service Provider is in an accident with some human driving like me, an old muscle car that hes renovated in his garage, and its still that dumb flesh and bones behind the wheel, well work these things out. It will certainly take time. We live in a very litigious society. Our lack of a loser pay rules will ultimately lead to a lot of these innovators sued because of deep pockets early on and discouraged from developing this type of technology. That does concern me greatly. But i dont know what we should do necessarily outside of changing the loser pay law. Good luck with that. But the reality is, im pretty confident that personal injury law will work itself out over time. And hopefully a lot fewer personal injuries. One thing is that unlike an ignition problem or Something Like that, or a sticky accelerator pedal, the Driverless Cars record everything thats happening around them all the time. So if theyre involved in an accident, they will have excellent 360degree recordings, what happened prior to the accident, and it will be fairly easy to determine who was at fault in the accident. If the manufacturer of the driverless car is at fault, they will pay up. And update their software. If they werent at fault, then that will be fairly easy to prove it. I think that will take care of a lot of the litigation problems and liability problems. Yeah, ill agree with both adam and randal. One interesting situation yeah, i think we should basically let the law evolve. Well deal with any of these potential problems when they arise. But one, you know, taking adams example a step further would be, what happens if youre in a nhtsa level three Automated Vehicle and you collide with a nhtsa level four, how do you start then, sort of breaking that out, and assigning liability to various parties. Its going to be interesting. But its something i think the courts hopefully will be well prepared to deal with. Thats the best option we have now, as opposed to pre emptive statutory changes. I think youll do more harm doing that than just waiting and seeing. When people ask two questions, i always forget the second question. Driverless cars will encourage more people to get to work and read and watch movies while theyre commuting. Why isnt it already happening. People who commute by transit tend to spend twice as long commuting than people by car. Thats according to census data. It is already happening. We have this question from the front, please. My name is gabriel roth. Thank you for a very interesting presentation. I understand that every selfdriving car has to know where it is at any time. Is it necessary for others to know where the car is . In other words, my question is, do selfdriving cars have to be tracked . Which is what i think i understood mr. Thierer to say. Im not sure i understand the question. Do they have to be tracked . Theyre going to be connected. And there are ways that they will be tracked in that sense. But maybe marc or randal, do you have an answer to that . I think the way that the at least what theyre telling us publicly, the way the current manufacturers envision this, they want to focus on developing sensor technology, so there will be a gps receiver. So the car will know where it is relative to other things. But thats what the sensors are doing. In terms of them being tracked, i dont think youll youll be able to pull up a screen and see where all the other cars in the vicinity are. But theyre going to you know, a selfdriving car is going to know where another car is. It wouldnt necessarily know where another selfdriving car is. I dont think thats necessary. Tracking is not a part of the technology that is being developed by any of the companies that i know about. Question at the back . Im kathleen sherman. Im from the university of virginia law school. I had a question about the accident avoidance algorithms, and if theres any indication from any other companies about the approach to this. Is it going to be, you know, every car for itself, for accident avoidance, just protect that car, or is it going to try and reduce injury to both cars, if reduce injury to both cars if theyre communicating or if theres any concerns with that. Thats a great question, as someone who studied philosophy in the old days, their socalled charlie problem that comes up in ethical philosophy, two charlies heading at each other, what do you do . You have all sorts of live and death trolly how to create more ethical algorithms, do you have two Driverless Cars hit each other. If one goes off the bridge, and the other is going to hit a car load of kids, which do you do . Through trial and error, we learn how to avoid that. No matter how thorny those ethical algorithms are, im confident that those intelligent vehicle technologies will help us avoid more death than when the dumb flesh and bag humans are doing, were going to make the worst of all decisions, like hit the bridge and hit the kids all at the same time. I can tell you this has resulted in some very heated debates, if you look at recent debates and wire patrick lynne, evan salinger, and others engaged in this. And theres a whole volume out there called robotic ethics where these questions are debated. I agree with adam, i think the main problem we have right now is that we have people driving cars. So and whether or not they can you know, i think theyre the ones who are causing the accidents, not to say that we should take it away from them, but this technology, theyre all engineering it, at least what theyre telling us these are going to be very, very cautious, so if they get into a really complex situation like that, chances are they probably already would have pulled over and stopped by the time they get to that point. They really are trying to avoid getting to these questions, these ethical dilemmas. But as you said, it is a really interesting question, i think in the fight its something that developers are going to be work on. We have a question from the gentleman on the left. Im from south washington, d. C. Also, this is a little bit different here, im a regular psy cycler, and most of us realize the advantage of cycling is sort of the asynchronous nature, most of us actually dont know about traffic lights and signs and stuff like that. So my question is for Something Like an autonomous motorcycle, i dont see this for a vehicle, is there any likelihood that they would have more flexible rules in the sense of you definitely can get places a lot quicker if youre allowed to sort of drive between cars, or in other words im trying to take this thing ten years beyond after its been introduced, but is there a chance . Because you can go, any time during rush hour, you can go by bicycle, anywhere near the white house and get from one end to the other in no time, but there is no vehicle short of the president , with all its flashing lights that can go anywhere in less than a half hour. In california you can lane split on a motorcycle. But an autonomous motorcycle is kind of taking the fun out of it all. I dont see there being a huge market for those. But, yeah, thats an interesting question, what you can do. I mean maybe, if some people do want that, you would talk about legalizing lane splitting going between cars, i guess thats, you know. I can see in the long run that things like stop signs and possibly even traffic lights, speed limits, things like that are going to be redundant, the car is going to look and see what kind of road its on, and whats a safe travel speed for itself and the need to have a law defining that safe travel speed is not going to be there anymore. But thats going to be an evolutionary thing, i mean you would say that this is a ways when we have, basically, were approaching 100 automated fleet. Can we take a question here, please . Thank you, chris moody from yahoo news, you touched on this just a second ago, if i wanted to drive faster and im in a big hurry, are the Companies Making options in the car where i can drive faster, if i wanted to save gas i can drive slower, also the implications for city Municipal Revenue where police now get a lot of their revenue from tickets, this will be a big problem, right . Where theres a sharp predict r shungation or theres no mechanism, also if they pull over a driverless car with dogs in, then what happens . Ill be worried about dog policemen. There are cities that do depend on revenue, ticket revenue for a lot of their revenue and thats something that theyre going to have to deal with. Theyre going to have to figure out an alternate source of revenue, its just like 70 somebody at a conference said that 70 of Organ Transplants come from auto accidentses, so should we ban Driverless Cars so we can continue to have organs for transplant, i dont think so, were going to have to figure out other technologies for that. I didnt mention in my remarks, we cant have this conversation in a vacuum just in talking about vehicles, you have to think about other technologies that can satisfy other types of demand. I spend a lot of my ive time thinking the cars, whether they be groceries or whatever else, theres a question can things be delivered vie tha flying robot. And that could mean more time in a car, Autonomous Vehicle, and a lot of things my children might use a robotic car for now, so we dont really know what kind of Butterfly Effects will unfold because of this Technology Plus other technologies that can satisfy demands we have. We have five minutes, so unless theres one question in the middle, right there, thank you. I know with the International Technology and trade associates. My question is, with Drone Technology and commercial drones, one of the biggest impediments is sense and avoid technology and i feel like sense and avoid would be able to develop with Autonomous Vehicles as well. I was wondering if you have seen any to develop this technology . I havent, the question is with regard to sense and avoid protocols with regard to some sort of commercial invoe nation, with a completely precautionary approach, that shalt not other may i is sort of the gain with any of the gain in this country. Lucki luckily, were seeing a what more fresh approach with the vehicle technologies. Well have to see, but as far as i know right now, no, theres not a lot of that happening. I have spoken to a gentleman who was working for a parts supplier, designing some of the Sensor Technologies and theyre interested in both the automotive and the aviation markets. Theres always some companies at least considering that. I worry about nitsa, but faa rooks like this nitsa i mean yeah, the right of way rules, the scene avoid requirements or the right of way requirements that the faa basically right now give the faa the authority to shut down anything i wants, and until those are revolved, yeah, its a big problem. Aerial drones is kind of a headline grabber, but the reality is, once Autonomous Vehicles are out there and able to deliver goods to people, i think the desire to use aerial drones is going to decline, since the weight problems are going to be significant. Its going to be easier to use a ground vehicle. A brings the q a section to a close, if you could make your way to the left up this spiral staircase, into the conference center, lunch will be served, the restrooms are to your right in that corridor, keep your eye out for the yellow wall. All thats left for me is to thank our panelists, please join me in thanking them. Next on American History tv, colonel gregory daddis. Professor daddis led to failures in the the new York MilitaryAffairs Symposium hosted this event. Its about two hours. Good evening. Its a pleasure to introduce our