Reporters' Roundtable 52: Self-driving cars!Two great experts who teach at Stanford talk about the coming robot uprising. Sorry, I mean when we're going to be sharing the road with autonomous vehicles.
-- -- -- One more time but that's about it -- technical problem. Hi everyone I'm Rick Adelman welcome to reporters roundtable our weekly dive into deep dive into a single tech topic of this week. Our topic itself driving cars. Our -- took of course is that is the -- New York Times story that unveiled the Google program to develop self driving cars. Cars which are already. On the California highway. Driving in traffic. Maybe alongside -- with people with Robert Scoble is surreptitiously -- to Google self driving car apparently when he was alongside the car makes all them. The point eight camera at him somebody hit the break now button and a car -- back and spent board. Anyway there have been other big stories in the development of self driving cars the first big news to get public's attention. Was -- running of the DARPA grand challenge for robot robot cars in 2004. A car built by Carnegie Mellon University won that year by driving the artist but it did not finish the course none of the cars -- in 2005. The grand challenge. The second year the grand challenge five vehicles finished the course and the winner with the system developed by stand by Stanford called Stanley. We're going to talk today about suffering cars and what's going on at Stanford and as the team there are preparing to take on even more challenges in the self -- -- space the trying to raissi Audi pikes peak for example. We have two great guests in the studio to talk about this topic -- -- biker in the executive director of the Center for Automotive Research at Stanford. Since spring 2009 he has -- the Stanford class the future of the automobile. A class I would love to take. He was -- BMW from 19952008. Working on technology scouting innovation management systems design and series development. You your car guy because they felt welcome thanks for coming in Sharma -- Also joining us from the future. Paul -- -- managing director of foresight at discern analytics. And a visiting scholar at Stanford Paul is a noted futurist whose work and essays have appeared in many many great places. Harvard Business Review fortune wired the LA times. Newsweek New York Times and Washington Post and -- in Pristina yet. Haven't we have to -- we're glad to have -- So. Let's get started where are we. Pardon the -- on the road to the autonomous. I think the first that is if you're on highway two -- and you happen ago. And see your car driving the speed limit -- it's actually a good bet it's a robot non human or my father. Seriously I mean having to buy cars are there other. I waited CME and I'm that -- print think pressure -- -- -- security need to figure out what kind of rolled -- -- -- -- along with the road because you probably wanna hear something I had just wrote it undermines long and -- the fifth minute when -- -- -- -- if I think upon it's actually the correction is what does this -- looked like and threw it to -- -- -- -- have been -- technology I think we just thought on this road and we are doing pretty well on the road but we're getting one while laying on the road and that we -- to figure -- these additional -- that coming actually on the road. Always gonna do with them and a lot off topic I guess that we should -- -- -- Scotland -- more laid off. Well will definitely get there are now as you say we are already. Making progress we what have -- where are we so far and what is this current. What robotics technologies to -- I -- -- term loosely are on the road with us right now. But it it's the only half right to a large and it will -- -- -- fifth and technology I'd on on private fifth and fifth and these all of these things you can maybe cause than analog ranking Fifth Amendment drive -- -- and fifth from piloted -- -- the driver. In breaking. More safely and and with maybe maybe -- -- And therefore if you will mean a quality computer on the on the Hudson the driver and it has some -- -- intelligent and now they. We -- on the -- called it and then robotic car and -- we do half being felt his thoughts. I'm am adaptive cruise control. The same thing compared curious than a cruise control now have the intelligence to detect that -- -- -- the com from a view. And slow down if neat Ian on the thing to Rio Ria was starting to feel the fifth and in the we are now entering the field off Madeleine controls though lane to lane departure what exactly -- he thought the thing. The did what we album we still on the field with -- drive is always in control the fifth and then -- office on -- vehicle. But with parking -- fifth and refute these things that you feel auto parking. Who really just opening the door and hopefully will not be huge pandora's box that we cannot manage anymore but -- the project I think he's from well let's. Probably framed this so we've got. Cars that can manage their speed and recruitment through laser or -- trying cruise control. We've got lane departure warning so -- -- -- -- -- on highway and we've gotten. Navigation systems the Google card knows not just where it is but how fast Ed tech I got a -- -- their 150 dollar GPS unit my card it tells me the speed limit and it's accurate most of the time. How big a gap is there between that and going on a freeway hitting drive and line -- -- if so -- parts lot of information and you're getting and everything -- think -- departure warning that really means you get a warning I but it does not really an -- with the steering system. And that would be really be a huge staff if you really hate the fifth and that to have take over control off your vehicle and an overall you feel whole -- of -- -- -- and then and then on. And then I think that might take quite some time receive fantastic research going on for all the right reasons then -- computer vision studies I'm. But to really have these the -- the fifth and control vehicle with -- driver. Fitting -- and then taking a -- -- might take over what's happened so far. Is -- gold has driven about a 140000. Miles with a couple probably about -- robotic cars. -- and a driver sitting there at their hands off him and a navigator on the right side watching the computers. So -- or autonomous but. With a lot of prep work. And then we had the third darker urban grand challenge -- in November 2007 where. We demonstrated that robots may be primitive but they in fact understand California vehicle code better than 95% California drivers. That was again on it a simulated. Urban settings and air force base. Set ups are like town with robots driving around and some very very humans driving in cars. And pretended to be pedestrians and so you have the existence proof that this can be done. But going from the existence proof to -- world in which we can go down to the dealership and purchaser robotic car. You know it's probably the second term -- the Jenna Bush Administration where you'll be able to do. That he never. Well I think I you know -- -- -- even Sebastian. Who who who is leads the effort at Google is still on record as saying. You know by 2030 half of all vehicle miles in the country -- The interest thing way to put it. After all vehicle miles implying that not all miles at least in my mind about a mile from the same that they highway -- -- answer their different -- gets deployed differential in the first place -- -- If I had place -- -- I think he could happen maybe by 2015. Moon is convoy vehicles for the military. The front where you have a lead vehicle driven by humans heavily armored. And then following -- behind it like elephants at a circus parade in robotic. Vehicles and and so you basically take the hostages out of convoy. And and it's and then if it's attacked the robots disperse and and you -- do some interesting things with that it was spends work has pointed out that. You know if you have vehicles driven by robots rather than humans they can safely tailgate so you could drive eighty miles an hour. With the robots you know ten feet off for the rear bumper the first -- so the military will have not only if you're hostages. But. Much greater flexibility in terms. Performance. The long term thing is the car that you -- -- be very expensive -- with more swap them but here's a wild card. Maybe this technology finds its way into car sharing. And and and you -- city car and and and and the car in the like that we we take the technology spend was talking about. The adaptive systems and we make not -- smart cars but brilliant cars. That are safer to drive and thus make a car sharing much more product -- And and we can get to it later written this is where I think Google's gonna play. While I do want to get to the cool thanks so let's let's get there why is Google. A information and -- advertising and search company -- doing autonomous vehicles. And this is not somebody side project this is happening at Google with Google dollars you want to speculate. Well well this is I'd I mean -- and I knew I couldn't speculated that just spent speculate again and again. Fill a book I mean it may be a good way -- cute try to answer the national believe we look at the automotive industry and what you can make automotive industry you've got manufacturer. -- those who make the caller you got supply enough -- to -- system and you've got infrastructure and service -- those who provide an update provide. The backbone for for many fifth and that's on Google and exactly so that the -- speculation maybe I'll actually -- well. Who -- in fairness really is important to note that Sebastian does take pains to say this is a research project they don't have any commercial -- but -- actually right. The action here is Google selling data. It's already you know if it's got Google Maps in it but -- -- day our busy mapping to greater levels of accuracy. They can sell that now the only problem they have is the automobile industries unbelievably conservative. Very slow to change that's why we have -- like spend at Stanford has to convince him to change so. You know they -- they may play at the edge of automotive -- addicts. This is where I think harsher it's. On so here's my scenario. You know Google already has a huge fleet of vehicles driving around doing the street view images. Unique on the Google campus that probably 200 people -- And and we have car sharing that exists but it doesn't quite work insurance company -- about it. If you put some of this technology into our shared car. Collision avoidance and -- -- you you protect against fender benders might lower the insurance and it might make it more convenient for people who only occasionally buy cars. Now I really -- volatile and our Google what I would also do is -- cameras into those cards -- and we would end up. Basically paying for the right to drive around in Google street view vehicles from. This is -- then they would just keep getting super -- -- if they could saturated city with enough cars here. Cars they would get real time street view and then they could start doing ads substitute and on the fronts of buildings and -- -- and other revenue stream and one more proof that it in fact is Google's world and we just -- -- -- I think there's some -- them -- them and because that also shows the logic then -- big veto that in fact if is that perverts -- thing by the the smart thing in in itself. But it's not really connected to the outside world and what Paul with -- rightfully and I'm just pointing out. I mean what the opportunities are -- really have -- car being connected to -- and infrastructure entered through backbone servers and whatnot. And how of the vehicle that can provide that data on all -- -- ought to use data and maybe this is just the -- a good example that we just saw with the vote Wii Fit on the new. I mean a scarcer vehicle like that is just a big sensor. And Google's a -- it's in the business of sucking up data and turning it -- to market. And and so. It does seem like -- somewhat expensive way to to invent a whole new field of science just seek new pictures of the -- about the you know what if there isn't famous comment made in the late eighteen hundreds. -- one of the fathers of electricity in sensor. What good is this for and he said I have no idea but she will tax one day. But a data Google has been done a very good job it during -- dated to step we can't -- yeah. The ballot -- -- -- so they are still many issues that yet to solve union that the drive button isn't quite there we talked with them things that that though autonomous vehicles can do the tuning in that -- called a caravan here listening. -- -- -- What are the challenges between here Andy you know Google self driving Prius for everybody. In the real world what what problems have yet. Yet India. What I felt I mean -- -- there's always one thing that it really comes up and these discussions and relief after five minutes at the -- that fell. -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- But just screwed up a little -- -- was nominated with a three items significance beauty are you gonna pay for the speeding ticket know you're only -- thick -- the card dated right. And -- -- -- -- and apply -- that we enact them on the thing flew in filled illegal. Implications are huge are unclear because there at least 3 am and keep it with all -- -- be manufacturer. Base the drive -- maybe than just to operate as a passenger on the com if the vehicle itself. And again might be okay and infrastructure provider who may be running them both to implement it that's one thing and infinitely -- -- the human aspects. Will people feel comfortable in the column like this -- we -- -- research going on at Stanford actually besides the engineering am aspect of it. How to people -- with an increasingly -- my automated because Phillips -- The because gonna break you might get scared -- white -- the break and what's going on here when -- a problem but -- because communicate with you. Hey look a few that -- come from a -- and therefore we need to slow down a little bit are you too much of lying to deliver but this is basically -- how few. That humans feel comfortable in the stink because after all the -- enjoy the ride -- that that the music and while. I want to get into that legal issue allowed because -- I'm fascinated by that problem there will be accidents. And Google has said look they're good they're Google cars and 14540000. Miles is only the one accident that's on -- one of our cars Margaret. But ultimately there's going to be a situation the software was not prepared for or -- flaw or something happens and somebody's going to get. This can be property damage somebody's gonna get killed it's going to happen has to happen. Who's responsible. And the person behind the immediate -- described by you know. Who is responsible. These are these are. These are wonderful. Topics to talk about over cocktails but this will get results so quickly. Once we have the cars we have a legal system that's very adaptable rooms we already have robotics psychiatrists there was one of them from 2004 grand challenge I'm just waiting. For the first robotic lawyer. -- -- you just have to get to a point where we can quantify the risk right now the risk. Button but I with you know you start getting miles on the road. The moment you can start quantify the risk then you can write insurance against it and you can create policy so the trust me you know weeks we sort this stuff out all the time the legal system won't. Robert who we will but don't forget I mean the automotive if you mention it uses very conservative and and also in some sense of -- -- industry with deploying new technology. Because you don't wanna huge recall. That you need to recall all of -- and even less so you are with you don't -- and it had line that fell that's -- company -- -- the phone company XYZ. Killed after -- if you don't want right because then you might go. -- -- -- maybe out of business oddly you what you would have among larger problem. Then then compensating for the bird magic off. And and I think the big correction and all of this -- what -- it actually mean meaning it the autonomous the -- And the question is what does it what -- it read like in the owner's manual. And it does it take something you need to think control at all times. Right even if it's an autonomous car with you in which case you're gonna have huge human factor -- to get to that effect and right and and and who's who's gonna pave that says that -- -- -- -- gonna pay an extra 101000 dollar phone autonomous car but you need to monitory at all times. -- -- -- -- -- If you made it a 100000 army but this -- doesn't -- you choose the bit. They'll apartment but we appointed actually makes a lot of sense because I am and often we get these comparisons. Where -- -- -- Right the we have auto pilot that can land actually airplanes in and pocket as a -- and everything faith and -- -- kind -- don't need to do anything. What did they actually do a whole -- -- monitoring it since but also they get paid. But also there is one major airline. It now has -- regulation on that the city unity the pilots are allowed to land the plane if they want. Except if conditions are below a certain level the pilots are not allowed to touch the controls to say. Basically pilot -- -- BS won't play around and -- a point but if the conditions are really debt. You know don't touch my. I don't wanna get off on a tangent here but -- -- you just that because this is a pet peeve of mine this delegates that Will Self flying airplanes they don't exist to to set up an auto land. In dense fog while it might be safer if it's appropriately set up. The pilots actually have to know exactly what they're doing to set up the -- -- -- the -- -- monitor and monitored carefully I'm just saying so. And and and related to close the loop here that's the actual receiving currently with the robot that we call them on parliament because. If you do half highly trained professional monitoring that -- these autonomous car wreck the -- what we are at may be -- -- -- professional and where where we on the wrong. Well also. This is not an issue for another reason that is you know thanks to work like -- -- -- student -- cars -- It's not that a robot it's gonna suddenly appear overnight what right now starting down a path is a continuum of adding more and more features to Carson -- time we have the first truly autonomous -- Everybody's gonna go -- -- you know. What took you so long it was about time permitted only to find what you're saying maybe they won't even notice because forgetting their bye -- inches for example anti lock -- as he talked about his driver rates now and -- to talk about a previous topic of insurance in fact you get a break on your insurance if your car has -- -- used to when it was uncommon you'd get a -- -- -- sorry for upon you get a break on your. Insurance for anti lock -- And other driver aids main pack lower insurance will the same thing happened we have cars with the drive. Button on them -- Yeah but I mean. It candidates and an outbreak this example is a very good warm enough to my knowledge knowledge in the beginning the insurance rates went down then but then they had -- and -- them again because -- something that risk compensation. The people who I have to -- -- breaks and I can redeem brightly -- -- because the break with hate me because they they do all of this what's going on. And and actually the cost more accidents on both of pretty that's a very interest in point of. -- you say about this in conversations -- we're we're having that paving the count past conversation where we mentioned. Cars are like cars except that now they have a robot but but this technology does is it really sentenced them to and and so the real surprises are going to be the ones we never think about like. The people are gonna be most annoyed by robotic cars -- -- where are the people who -- The car drivers are like today years you annoyed if you don't drive a Prius you can be annoyed by Prius driver because they're sitting there being sensible and saving -- -- and everybody else is roaring around so you know you'd have these cars mixed in with the rest of us and actually update you book and people are gonna hate that. But the real opportunity here -- you know is -- noted auto industry's very conservative. What this technology does plus electric motors is create an opening. Or some new challengers amid an out and nowhere in Broadcom the next -- -- our company and I don't think it's gonna be Detroit. And if you think about robotic control makes a lot more sense. When you get an electric engine and you get the stupid internal combustion engine out of there as you can if you when you're going all electric -- much greater control. And so I think this is an opening for. You know by the Chinese company. To come in and created nude semi quasi -- electric car. It's fundamentally if you were but Tesla wonder there were on this but we don't test as a -- company it's it's it's not car company to try -- company. The -- of it -- carton had wrapped car around to try entering in order to -- to try train early adopters but really what they're -- businesses like licensing it so. I am sure that Tesla is looking closely this. So are we gonna need different highways are gonna have robot only lanes and and we've got the diamond lanes and have are two -- or something. The -- that that quite a few things that we are looking at pictures if you what could be yeah I'm I'm an entry scenario for the fun of technology in and therefore specific lanes that a clothes for all of them because might make a lot of -- Although from the -- you have enough -- with a human and block off the land on the thing. And different highways. And she pretty interesting if you look back at the at the history off autonomous driving which both -- 23 decades ago caused. Automated highways. Because he had a high integration of the veto with a highway infrastructure. Actually our colleagues on the other title that I participated -- -- and Brooklyn it is great resource really with magnetic. Manned units and in the wrote surface and then there -- tents the from the car and they could actually Trace well whether whether lane will be by means of these -- throughout the that's an -- -- different way to do autonomous vehicles but it requires a lot of interest our infrastructure -- with the -- project and I believe the -- projects are about using existing infrastructure -- smarter and smarter cars and an infrastructure -- forever read the only an infrastructure of if you -- -- -- GPS you can call this and the -- -- you the magnetic that you don't have to read from -- Anything involving a century -- we don't we don't to infrastructure anymore in the country. It's very offensive -- I mean you know we can't -- -- bridges. So it's a safe bet that. The thrust on this is going to be. A communications link with the article and using exists -- existing infrastructure like satellites but. If it requires doing stuff on the roads. Then it yet had several decades -- exactly now that we don't we're probably not -- Indeed it's a sensible way to do things but you know. Look at I think that this is really what they're whole field in and then in recent shifted from automated high risk to autonomous -- The -- you cannot wait for the infrastructure. And another little hikes haven't actually August because you because infrastructure. Which would be a whole another discussion marijuana and but it's not happening because can be infrastructure -- department of transportation's would -- behind this topic. The union the on the Stan if you need to mutual what -- -- really -- the deployment of one non. You -- waited for over a long time and also the -- With infrastructures. You can update things if if you put the interest if you put the smarts into the thing that has replaced more frequently come and you automatically update if you embedded in concrete it's their for a long time. It's why nobody buys a cellphone built in their car anymore and -- -- -- -- cell phones are fashion your place of every nine months. And you replace your car every couple years if you have a built you know and electronics on your car you're gonna be stuck with obsolete -- Right right yet people are still spending 2000 dollars for navigation systems 150 -- -- over the top as well but. Maybe if -- point that's salesman. And what that does that fit when and if Hispanics again find him most people who want to map -- -- from the car and they just look at. They do now what I've started -- -- iPhantom -- -- -- a little bit. I want to talk about this we mentioned briefly about the government in building infrastructure and how many of these projects became famous the DARPA grand challenge. What is the role of the challenge of the -- rates. The the contest in the development of autonomous vehicle technology. Well. The fact is it put autonomous vehicles on the -- mean you look at. -- -- department -- defense did the grand challenge to begin with because they spent several decades funding autonomous vehicle research. And the results were absolutely -- that that the defense contractors who did a terrible job and so. They said well let's do this getaway special will -- million bucks and let anybody come. Anand and what you have if you look at the sequence of the three current challenges. May 2004. About 21 vehicles only by vehicles got any distance sandstorm lead vehicle died seven climbers. It was pathetic. Then almost exactly eighteen months later one -- period of -- law you have the second grand -- Or 22 out -- 23 entrants got farther than the than -- -- the first race. -- car's finish and and by the way for the record -- did not -- the first race nobody won the frustration Stanford on the sectors and courier Sampras pursuant. So you start seeing that -- remorse on then you have a third grand -- -- ground state urban one. Much steeper task -- and huge success so. This we're pictured here is that Stanley from -- 2005 -- count right there right now. Now the downside to this is because of that outlandish success. The grand challenge. In everybody now thinks that the way to do animation -- to have challenges and reassembling the normally like different rate. Peter do you -- this is running X prizes which are all great their central but these are not a substitute for research. These challenges. Work when you have a whole bunch of research that's been built over decades. It's that needs just a little -- to bring it all together into -- to station into realization. It's not a substitute for research it's it's instead -- research and development its search and redevelopment. And can it can be that catalyst in the way that you know you can have a a catalyst that causes of pearl the grown and oyster. It's interesting because many of these challenges in the expresses great example that as well with the prize. It and I don't believe. Offsets much of the development costs I don't believe that -- him you're the Stanford. Programs -- there are several professors and many students working on these projects. If they win the prize that. The money makes that they -- different there's something else at work here yet. What I mean come competition policy is something that that attracts just humans do to compete with one another and and and show what would -- comments on him -- another of for a good example of what's called the -- to had a man on the moon might. I mean what would we have fallen in on on the moon and -- nine without competition from I don't know. Right and and so it if something and you can do great things and -- you agree with Paul. You need the -- it really treated to two world and all these issues to really understand it for a while before you actually do something. But then. It really think that then added that a challenge a competition when everyone called it -- and -- finally together to say you what next October it is. When you -- -- or -- -- but we wanna see what's possible and insulin like okay dude and this camera no October. It's a wonderful way to harvest the research gotten. And and by the way students like it I can tell you that actually like it felt it at the great thing cute cute really push and one of -- -- more on these these projects really don't make -- want to go back to school. Actually commitments there when you finally arrived at my income next winter -- psychology degree night and acted too much and engineering I teach forecasting at Stanford engineering school I would like to have become into the class and we actually have a very great researcher in the psychology department looking at the human. Beaten and action that is one of the questions -- -- we have a couple more from readers here I -- -- -- -- get to them but just. -- briefly sure. When you're sitting in a car and it's driving itself I mean like you think who's gonna pay 101000 dollars or more or less or whatever and to have a car that they have to pay close that they have to worry about being organ. Because they're still I can't I can't imagine that we're not going to have the human control at some point again -- -- it -- it back probably to the fine print it just by the time point nine. If you allowed -- to beat texting in this car. Wouldn't you be paying ten -- -- for the front yeah and if you allow -- to -- -- the all whatever. You -- take a -- I mean that the very -- of scenario but if you can do other things. You might be very well Wilmington to pay and -- on little excellent on. But then -- responsible in a car that's arguably -- hundred miles an -- ten inches off somebody's -- I'm somehow responsible for the what -- a -- did this is exactly that it's just -- -- -- And this is what Amanda audio by independent world was because its orbit the side -- and the autonomous car isn't really autonomous or is -- -- -- -- that -- human being monitored. And and also it may be a vehicle that. Has different -- -- on -- on -- five well and on on five the cars driving itself. And then in more complex situations. You takeover even though rationally probably severed at the robot drive in them so you'll have different modes just like -- interest and. Now there are developments that are ongoing I don't know how much we can get into the nitty gritty what's happening it's -- -- in particular one of the that the big project right now as I understand its effort is the car and -- it's being designed to race up pikes peak correct which. This is hill climb this is a a climb -- -- winding mountain road on gravel. Where that goal as I understand this correctly is to have the cargo flat out and drift its way up the hill just like professional drivers do what are some of the challenges that are being worked through here. Well it if you -- maybe look at it from an -- how's it different from. What. We Fiore was -- arm and Brandon urban challenges the vote I'm honestly these them completely autonomous -- there were able to navigate an -- -- on the death of a film like that. And what this is what we saw there -- from professor Chris -- whose who's conducting research on the autonomous Audi -- at them. Is is really am. How can you control the -- at the limit off off handling often -- off adhesion. Basically what we know today as because stability control -- them. Right it makes you can stabilize the because but it don't have any notion of whether wrote -- -- might actually be a very important question to ask if you rather wonder. Compromised but if you little bit here but -- keep it because on the road and then once you get it on the road you can stabilize -- again. The what they're doing there is really tear it to see how the -- killed. Can be controlled at the limits -- off off its performance if you Wilton which was not the goal off that topic when an -- challenge -- -- -- with more led to avoid obstacles any thought of while object on the road. Which this -- such as not doing. And and also this is not necessarily something where you would say okay you would really have an autonomous the goals and racing stifle file powerful from like this. Which to adopt a -- -- -- anything basically if the patent. -- with autonomous car. No this is -- -- something really helped convict because if fifth private -- and very critical situation such as when it's about to spin out. I think and there were venue ought to need to look at how do you do this hand over between driver and because it looks what's what they researching -- on. Really okay when the become announces to the drive and a way okay you know what I need now -- and you can get you back on the -- -- When a vehicle announces to the driver. I need to help fuel the vehicle that's the driver I need your help and I guess we're both -- gift -- a pent up well. That is exactly what a mention of audio that that's not the path of that we have analysts and their ethic of -- Who exactly we researchers if communication between private and people that's thanks -- that if we've got a note. Call from Ontario says how -- -- council snow and -- -- to the car that's capable self driving and dry weather for example stopping time can be quite -- on and I wrote I guess this part of the question that we're being -- -- -- in -- -- because exactly this -- that's exactly what it trying to address from. And despite the recent. Appointment if that. -- the indeed the answer. While and also -- is where the that DARPA grand challenges are informative and humans versus Hickel. One of those wonderful coincidence was that thirty minutes before the November 2007 urban grand challenge. 250 miles to the north on highway 99 there was a 108. Vehicle pileup and fog one -- thick fog oddity that bug it's always been there everybody knows it's there and people like salmon you know they just instinctively. Keep going haven't yet either eighteen big -- in this thing you know the firemen who were rescuing people from the front heard educate your car smashing into the back. -- if ever we had proof. That humans shouldn't drive. This was -- you know -- -- maybe bring it the same day that we saw robots primitive robots have been the vehicle code. You know we saw Californians too stupid to know to put on the brakes but so. But but interestingly enough and that's another that the statistical rather an opinion poll that 70% of all a driver thing a day are about average. Fined 70% I mean that's optional possible right and knowledge of any 5% thing that I average driver so people we have -- -- -- -- -- it's heavily skewed towards the male side is probably you know 100% of male -- -- that women are more sensible than a week we did we reported. Months ago about the study that shows that most people think they are very good multi tasker and -- 3% of the population is able to united. Have their attention -- more than one -- at a time everything but that's exactly why people namely all of weaponry like I'm not but need more assistance from the vehicle. So which brings up another question from -- beast that's how do we do -- that the people will be selling these -- plan to get consumers over their inherent distrust for computer controlled anything. And I I'll -- that -- is -- the worse the toward the environment the situation -- that the cards and for example a thick -- want to be the less likely that people are going to trust their computers when it is when they need them the most. Again this is going to be a gradual process and that it's not like we're going from the model T -- the -- -- We're gonna be just adding more and more features so it's -- continuum -- People won't notice. Also the other thing two years. I I -- Really have my doubts about whether we will ever actually sell a robotic -- Because when you add in the change in power plant from from the internal combustion engine from ice to electric. And then all of this rising -- -- on the control. It may look like a carbon is not card. And so and and the technology update. Already the the -- percent of and automobiles. Value. Is that based on software you know depend on how you measure -- more than half the cars values in the software already. I think we may end up in a world where we no longer buy cars by the time these are right we subscribe to cars. Right that's what better place trying to that are -- -- -- -- -- -- at first but it can take things like better place and -- cars and -- -- And this is like the early days of the -- purposes in -- -- was around. In the in the first half the last century they sold -- they rented old cars from downtown and it was Avis he had the big idea noticed that airplanes were coming. And airports is that wait a second I can reinvent this thing -- so his idea what I'm gonna -- my cars at the airport. And only wealthy businessmen -- traveling by plane so it's an upscale customers -- the go to Detroit. And say you guys are gonna give me a discount on your newest cars because in essence this is people paying for the privilege -- test driving your cars. Reinvented the whole business and I think long before robots ride robotic -- -- right -- so fundamentally reinvent this business that. The idea of buying a car will seem very straight it's already in fact happening you go to Germany you go to Japan the same thing happening here. There did that young people in their twenties don't want to own -- Now users like us people -- -- -- that was our -- astonished when. Twenty somethings don't want cars so the whole car culture as disappearing as quickly as the automobile it is. That I would love to end -- that there's one more question which we half -- Because -- for the for the users like me who still driving their cars are a driver thirteen year old car what about potholes. What can robots -- to save our Carson destroying themselves in our city streets. All I mean film festival -- they can actually track August pothole might mean they have these great -- the stands in front of the vehicle they can now I'm avoid them pretty impossible they they can slow down they -- -- but yeah I mean you yeah are you using. Service radar and and -- detect in the -- before -- And even more so I mean did just just think again infrastructure they can upload all this data you know what I'm seeing something infected his growth segment and then -- -- -- another car comes along and thief it's growing. We can actually predict these things. And then there'll be a robotic. You know asphalt truck that comes along and it'll have too -- -- pension with the state as well oh we're getting carried. OK. Well that wonderful note. Then thank you very much very welcome where can people find who thought that you're working on. That it actually had automotive -- that but don't EU great thank you spent and Paul. Well that -- again people wanna read more of your stuff where you recommend ago. -- her dot com come to my sites -- And -- -- piece I wrote in the -- piece of the search -- -- -- -- -- -- Often vote stuff well that's what it. You know what I mean I -- a robot to do this I guys thank you so much for coming in thanks all for watching us would back next week if you want to know more I get all the links to the show. And all the stuff we talked about its -- reporters roundtable dot cnet.com you can send email to the show roundtable at cnet.com and while Omnia Twitter I'm -- are a FE. In the volume. Twitter. -- -- -- -- -- -- Not -- I am button. -- might not -- I I I think you prefer email which is by -- the mine my last name at that but you okay. Great -- and gentlemen thank you very much -- back next week thanks everyone have a great we --