Home / General / Self-Driving Cars

Self-Driving Cars

Comments
/
/
/
950 Views

la-20140528-003

Atrios is right. Self-driving vehicles ain’t going to happen, at least not in the United States.

Volvo’s North American CEO, Lex Kerssemakers, lost his cool as the automaker’s semi-autonomous prototype sporadically refused to drive itself during a press event at the Los Angeles Auto Show.

“It can’t find the lane markings!” Kerssemakers griped to Mayor Eric Garcetti, who was at the wheel. “You need to paint the bloody roads here!”

Shoddy infrastructure has become a roadblock to the development of self-driving cars, vexing engineers and adding time and cost. Poor markings and uneven signage on the 3 million miles of paved roads in the United States are forcing automakers to develop more sophisticated sensors and maps to compensate, industry executives say.

And I am glad they won’t happen either, as trucking is one of the last OK paying job for working people without college educations.

FacebookTwitterGoogle+Share
  • Facebook
  • Twitter
  • Google+
  • Linkedin
  • Pinterest
  • Ahenobarbus

    Bad news for the 35,000 a year who die in car crashes. But I don’t think that some difficulties with the technology at a press event quite proves what Erik thinks it does.

    • I like the underlying assumption that technology won’t fail.

      And I assume you aren’t reading the full article since it goes on to discuss that terrible state of American roads causing major problems.

      • I’ve had enough autopilots go “stupid” on me to know that it can indeed fail and most unpredictably.

        • tsam

          Whoa–what does that mean, “go stupid”? (Remember I fly on commercial aircraft occasionally 0_o

          • N__B

            Obviously, the Major flies on planes that intend to vote Trump.

            • tsam

              Holy terrifying.

            • Trump does have a 757.

              • tsam

                Let’s see how he handles a stall then.

                • jim, some guy in iowa

                  with a wide stance no doubt

          • Here you go. I wrote about one such incident a while back.

            http://www.dailykos.com/story/2015/02/11/1363619/-When-airliners-attack

            • tsam

              The engineers were thoughtful enough to put a little yellow indicator on the airspeed tape that shows our stall speed. Right now the yellow “hook” is sitting right at 200 knots and the airspeed is somewhere between 240 and 220.

              Oh. My. God.

              So if you hit a stall at that altitude, full of cargo, is it possible to recover?

              • You could recover but you would lose a lot of altitude in the process.

                Also a stall in a heavy jet can actually cause structural damage. We only practice stall recovery in the simulator for obvious reasons.

                • tsam

                  Yeah, I kinda figured free falling in a heavy jet wasn’t very good for them. That would have scared all the living shit right out of me.

            • Ahuitzotl

              that was fascinating, thanks

        • Thirtyish

          That was basically what happened with the Air France crash in 2009. The autopilot stopped working, and the pilots had no goddamn idea what was going on or what to do.

      • delazeur

        It’s naive for you to think that this is a question of if self-driving vehicles will take over in the United States. For better or worse, it’s a question of when; our poor infrastructure is only going to slow it down.

        It’s also wrong to talk about whether or not the technology will fail rather than whether the failure rate is larger or smaller than that for human drivers.

        • urd

          Why is it just a question of when? Please prove to me why this is going to happen.

          Hoping something will be the case is not proof.

          • delazeur

            Obviously I can’t prove that it will happen any more than you can prove it won’t, but it is absurd to claim that “poor infrastructure” is an insurmountable obstacle given the amount of resources being dedicated to creating this technology. I’m not sure where you’re getting the idea that I am looking forward to it.

            • urd

              Actually I think I can prove it won’t: look at where the world and the environment in general is heading. Within our lifetime projects such as this will be considered a waste of resources as most key technological and urban engineering projects will be focused on keeping the water out of our cities, moving the cities to higher ground, finding new sources of clean water, and all the things that will need to be addressed with massive sea level rise.

              While there may be driver-less cars here or there, they are not the wave of the future for the simple reason that there will be no excess resources for such a frivolous endevour.

              As for the “looking forward to it” idea, I apologize, my mis-read of your post.

      • leftwingfox

        The technology doesn’t have to be “perfect” to be successful, it just needs to be “better”.

        A self driving option that “only” has a 50% reduction in crashes and fatalities, while allowing transport trucks to run all night without sleep?

        Boom, economically feasible.

        • John Revolta

          This.

          “I don’t have to run faster than the bear, I just have to run faster than you.”

        • ColBatGuano

          Who is liable in a crash then?

          • liberalrob

            The car owner, I’m sure. It’s up to them to sue the car maker. Gotta have more lawsuits to employ all those lawyers that keep getting churned out!

    • BGinCHI

      Good news for Abbie Wambach.

      What? Too soon?

      • Dilan Esper

        DUI’s are another good point, actually. This could solve that huge social problem.

        • Rob in CT

          That’s pretty much built in when we cite casualty statistics, though.

        • urd

          Actually DUI’s are not a good point. Most states are going to require that a driver be able to take over from the car in the case of an emergency, like an OS failure.

          If anything it is going to make things more complex as states may also require a system to determine if you are drunk or not before you even start the car. So while it might stop drunk drivers from getting behind the wheel; it is also equally likely to kill enthusiasm for the whole concept.

          • Dilan Esper

            What do you think is going to happen to the accident, injury, and death rates associated with drunk driving?

            I mean, in the threads on college drinking, I ran into plenty of sentiment here from people who argued that any attempt to use the legal system to reduce the drinking rate is the same as Prohibition. That isn’t how I view the issue, but if you do view things this way, it seems to me that self-driving cars could be a very good thing in any effort to allow people to get drunk with fewer externalities.

            • urd

              Did you not read what I posted?

              A self driving car, by itself, is not going to change anything because states will enforce laws making sure the driver can take over in case of a failure. To prevent people from gaming the system and using their self driving car while drunk anyway, they will pass laws to install some type of system to determine if the driver is drunk. If the driver is drunk, the car won’t start.

              So for that whole section of people that want such a driver-less car because they can go out, get drunk, and then be driven home – the car is now useless. They are better off having an older car that doesn’t have such technology and taking their chances. Most people who would be okay with a car “telling” them when they can and can’t drive are not likely to be the people who drive drunk anyway.

              • Ahuitzotl

                A self driving car, by itself, is not going to change anything because states will enforce laws making sure the driver can take over in case of a failure. To prevent people from gaming the system and using their self driving car while drunk anyway, they will pass laws to install some type of system to determine if the driver is drunk. If the driver is drunk, the car won’t start.

                Your ability to foresee the exact acts of a bunch of crazy republican state governments is impressive in its total fraudulence.

    • mds

      Bad news for the 35,000 a year who die in car crashes.

      You know what else is bad news for them? The inadequate availability and funding for already-developed mass-transit options for many of them, while rich Silicon Valley fucks chase a currently-unproven technology that’s primarily for their own benefit.

      • Ahenobarbus

        I’m sure if we just stop self-driving cars dead in their tracks, America will wholly embrace mass transit.

        • witlesschum

          No one argued that, but whatever you need to tell yourself.

          And Americans seem quite willing to embrace mass transit when they have access to it and it works.

          • Ahenobarbus

            And Americans seem quite willing to embrace mass transit when they have access to it and it works.

            Only if you have a selective definition of “Americans,” but whatever you need to tell yourself.

            • DrS

              Yes, it’s true. No one rides the subway in NYC. BART in the SF bay area is hardly used. The issues with the DC Metro are no problem at all cause no one rides it, so shutdowns will not cause any problems.

              Seems like you’ve got some pretty selective definitions of Americans, bub.

              • Thirtyish

                This.

                • urd

                  Just to make sure, this is snark, correct?

                  I can’t speak to NYC but in the Bay Area BART, and CalTrain are HEAVILY used to the point the systems can’t keep up with peak demand and both are having to find ways to increase capacity.

                • Thirtyish

                  Yes, as someone who daily uses the MTA, I can assure you that it’s snark, riffing on Ahenobarbus’ incredibly inane anti-urbanist comments.

              • Scott Lemieux

                Only exurban white people, preferably in the South, are Americans. Get with the program!

            • witlesschum

              I defined it right there as those who “have access to it and it works” which indeed explained that I was selective. Because I actually try to post things that are true, rather than whatever game you’re playing.

              • brugroffil

                What you’re saying seems really circular, though. Mass transit is going to be built where people embrace it, and people with access to decent mass transit embrace it. That doesn’t mean that mass transit is always the practical option or would always be embraced. I’m also struggling with how “but mass transit!” is some sort of a critique against auto manufacturers and technology companies trying to build self-driving cars. It’s not an either/or.

                • witlesschum

                  I also don’t think it is an either/or with self-driving cars, though I know which I think would be a more productive use of resources.

                  I was just responding to the statement made. I realize it’s kinda circular, but I think the fact that there’s currently not as many places with good mass transit as people demand definitely means it’s bullshit to say American don’t demand mass transit.

          • The Dark God of Time

            We now have 6 bus routes in my town, 35 years age that number was zero. And we’re the Gateway to the Gateway to the Sequoias, smack dab in Red California.

        • Ahuitzotl

          well if you stop enough of them, the roads will be unusable, I guess …

      • njorl

        Self-driving cars will make public transportation more attractive. When you separate the “driving experience” from transportation, you help kill the personally owned vehicle. The natural evolution will be a continuum of vehicles from the driverless equivalent of taxis to vans up to buses. It will also mean fewer vehicles on the road, which will also put people out of work from good paying jobs in automotive factories, but the benefits should be made to outweigh the drawbacks.

        Anything which helps kill the requirement of car ownership for participation in society is generally good for a variety of reasons.

        • Chuchundra

          I think you have it backwards.

          A personal vehicle lets you plan your travel on your own schedule and travel via the most efficient/direct route. With some exceptions, it’s generally the least time-consuming way to get from point A to point B. This is its major advantage. Not the driving experience.

          While I do enjoy driving, the times when I generally take mass transit, traveling into Manhattan from my home on Long Island, are when the driving experience would be very bad.

          If the car could drive itself, I wouldn’t care so much about that and would happily read or nap or watch a movie while I sat in traffic.

          • Rob in CT

            With some exceptions, it’s generally the least time-consuming way to get from point A to point B. This is its major advantage. Not the driving experience.

            Agree. I really don’t think love of driving is keeping many people from using mass transit. A few, perhaps, but mostly it’s about (real or perceived) convenience.

            • witlesschum

              Where that breaks down for me is parking, which is just a pain in the ass, even in a relatively car-friendly place my car feels like an encumbrance a lot of times, compared to stepping off a train or a bus or riding a bike.

              • liberalrob

                Well sure, if I could step off a train or a bus in less than twice the time it took me to drive 15 miles to the office, I’d be doing it. The problem is, I have to walk three blocks to the bus stop (10 minutes), wait for the next bus to show up (avg. 15 minutes), ride it to the train station (5 minutes), get off the bus and go to the train (5 minutes), wait for the next train (avg. 15 minutes), ride the train downtown (30 minutes), get off the train and go to another bus (5 minutes), wait for the bus to the office (avg. 15 minutes), ride the bus to the office (15 minutes). So it’s a ballpark 2 hour odyssey to use mass transit to get to work for me, or a 30 minute or so drive. Which one am I gonna pick.

            • gmoot

              Could be love of driving is one reason why so few people buy fuel-efficient cars.

              • liberalrob

                Could also be the lack of fuel-efficient cars. Or of fuel-efficient cars that aren’t crap.

                • weirdnoise

                  It’s “safety” or “masculinity,” both of which require a lot of rolling steel to propel around fuel-inefficiently.

          • njorl

            Don’t think of public transportation as just mass transit.
            Assume reliable driverless cars are a reality. You will be able to buy the transportation you want instead of the car. If nobody in your area wants to buy the same trip as you, you’re stuck buying the most expensive trip. If a few people want the same trip, or parts of the same trip, you can buy a van ride for less cost. If part of your trip is on a main thoroughfare, several vans link up to travel more efficiently for parts of your trip and you save a little more. Instead of bus routes and taxi rides, you have a full continuum of ride choices.

            If you don’t have to get behind the wheel yourself, if you don’t have to worry about how it handles in the snow, if you don’t care about the status it imparts to you, you won’t want to own the car. You’ll just want the ride.

            • liberalrob

              if you don’t have to worry about how it handles in the snow

              I eagerly await the technological solution to my car sliding down an icy hill. What does the driverless car do when George can’t reliably control the car? Sometimes wildly trying to reestablish control (say, by slamming on the brakes) can make things worse…

              I’ve wondered what happens when an idiot pigeon flies in front of my front bumper (which has happened more than once), or when my driverless car rolls into a cloud of lovebugs, as we often get in the spring here in the sunny South. Will my collision-avoidance radar be slamming on my brakes every 5 miles?

              Driverless cars are one of those Buck Rogers ideas that sound great in principle but are an unending nightmare of issues in practice.

              • so-in-so

                There are anti-collision systems in use right now. Do they slam on the brakes for pigeons or clouds of love bugs? Maybe they do, yet they are being sold and are in use, so it isn’t like this is an aspect of the technology that needs to be speculation.

    • Coastsider

      Actually, if people would just use existing technology (a seat belt), this number would almost cut in half. Unfortunately, technology can’t save people from their own idiocy: NHTSA Fatality Analysis Reporting System (FARS) Encyclopedia – People: Restraints

      • witlesschum

        I’d think you actually could pretty easily create a car that couldn’t put into drive or first gear without a seat belt being buckled.

        • There was a brief period in the early 70s where cars had an interlock that wouldn’t allow it to start unless the seat belt was fastened.

          People quickly figured out how to bypass it.

          • witlesschum

            Huh, never heard of that.

            • My father had a 1973 or 1974 Ford Station wagon that had this feature. It also had a switch that sensed when someone was sitting in the driver’s seat.

              By lifting your weight off the seat, the car could be started without the seatbelt.

              • liberalrob

                Any safety feature can be circumvented. Any annoying safety feature will be circumvented.

          • jmack

            Yep. My ’74 Cutlass had a sensor in the seat that forced the driver to wear a seat belt or the car wouldn’t start. It often malfunctioned and I disconnected the sensor.

            • I’m shocked that a 70s American car would suffer frequent malfunctions.

              • ColBatGuano

                I never had any problems with my Pinto before it burst into flames.

        • creature

          That was tried in the ’70’s, it didn’t work out. Most of the systems got disconnected. The technology was pretty rudimentary, so, this new-fangled stuff could change that. The new GM cars have a setup where you can’t turn on the radio without fastening the belts. You know, so the kids won’t drive without a belt. Perhaps if they started texting, the car would pull over and stop, that might be useful.

          • witlesschum

            I’d think it’d be more accepted now, in the world of seat belt laws, and buckle up public service announcements, but that also might mean the people not wearing their belts are the dedicated who would find ways around the technology, even if it was better than the 70s technology.

            • so-in-so

              I think the other “old school” work around was buckling the seat belt and stuffing it into the crack between the cushion and the back of the seat – there, always buckled!

              I’ve lost track of the number of people who know of the friend of a cousin who only lived because he was thrown clear of the car before it burned, went over the cliff, the dump truck hit it… never, of course, direct knowledge of such a person. I find I’ve buckled my seat belt when I move the car down the drive way – its just a habit after so many years.

              • brugroffil

                This actually happened to my grandpa! It was some time in the 50’s and he was stone-drunk. He was driving his convertible, passed out, fell over into the passenger seat and then slammed right under a semi trailer. Doesn’t stop him from regularly buckling his seat belt these days.

              • witlesschum

                Yeah, that seems like the first one I’d try if I was so inclined.

                See also motorcyclists who are convinced that they’re better off with helmets because they think they can more effectively see trouble coming.

              • Rugosa

                I was in one of those accidents where I probably would have been killed if I hadn’t been wearing a seat belt. It was about 1973, the car didn’t have a shoulder belt. The vivid memory I will take to my grave is the feeling of the lap belt holding me back as my head hit the dashboard. Without the lap belt, I would have gone through the dash. As it is, all I have is a barely noticeable scar. You can bet I buckle my seat belt every time. And tell this story any time someone brings out one of those my cousin’s friend stories.

                • steverinoCT

                  My otherwise sensible neighbor never wore a seatbelt when I was growing up (don’t know about now that she’s elderly). Had a close relative killed when hit from behind at such an angle that the belt crushed her.

                  Pointing out the odds of that vs. almost any other situation was useless. I myself got in the habit with my first car that had a very slick bench seat. As long as you didn’t make a turn you were fine.

                • I flipped a car at 55. The sensation of hanging from my seat belt/shoulder strap is still vivid.

                • Jhoosier

                  I got T-boned on my side of the car, caught between the collapsing door frame and the seat belt, my pelvis fractured in 3 places.

                  Considering the car also flipped upside down, I think I came off better than if I hadn’t been wearing one.

                • liberalrob
              • Ahuitzotl

                it’s amazing how many people know that guy!
                .
                I was seriously freaked when I first arrived in Arkansas and they were debating a compulsory seatbelt law, with all these loons insisting it’d result in you getting trapped in your car and burnt to death, in any accident. That was 2008.

                • so-in-so

                  Too many ’60’s TV shows, where the cars burst into flame (and if its the good, you see him get up off the pavement after the explosion).

  • jmauro

    What you are likely to see though is automation in the act of driving like auto-breaking or auto parallel parking.

    Things that are pretty self-contained and easy to automate, but not a “sexy” or “science-fictiony” as a completely self driving car, but still really useful for an average driver.

    • rea

      auto-breaking

      My car broke itself.

      • jmauro

        Sorry, auto-braking. Auto-breaking has been standard since the Model-T.

    • Brett

      That’s already happening, and some of it is even more advanced in some of the Tesla cars.

      I’ve said it before, but you’re not going to see fully driverless cars on public roads in the next 10-15 years aside from prototypes (at least in the US – Europe’s farther along on some of this and testing driverless trucks). What you will see is “driver-operated” cars where more and more of the driving is done by automatic systems, to the point where 95% of the time the driver is there simply for liability reasons.

  • keta

    This might seem like a dumb question, but why do so many people think we need self-driving cars? Seriously. Is this something we’re all clamoring for? Proof-of-concept technology run rampant? A combination of the two?

    • Some justify it by claiming they will be safer. I am skeptical. But I think fundamentally it’s America’s technological fetishism at work.

      • Dilan Esper

        Car crashes are one of the biggest killers. For some reason, a lot of people, including you apparently, think it is just fine that 10’s of thousands of Americans die when we could make the roads safer.

        Meanwhile ISIL kills a few people and Ameriva gets its knickers in a twist.

        • keta

          Jesus. Laughable assumption that then veers into total non sequitir.

          Too bad technology isn’t developing auto-pilot critical thinking skills, Dilan. Make discussion boards safe, and all.

          • Too bad technology isn’t developing auto-pilot critical thinking skills, Dilan. Make discussion boards safe, and all.

            As I’ve stated before, there are many problems technology will never solve.

          • Dilan Esper

            So all the scientists who think these technologies will reduce the death rate are wrong? Why, exactly?

            • tsam

              For the same reason a bunch of these newer miracle drugs kill people. There’s a process to developing technology, and these guys are trying to ram this one past the development stage–knowing full well that the current state of infrastructure makes the technology unsafe. (Or at least not any safer than the current tech).

              The scientific method. Test–replicate–review, then conclude. They’re trying to half-ass the first two steps.

              • Dilan Esper

                That may be true, but it isn’t Erik’s argument.

              • Pseudonym

                There’s a big difference between saying technology development needs safety precautions and checks and balances and saying patients would be better off without COX-2 inhibitors (especially because they might put well-paid chiropractors out of work).

                • tsam

                  No, but there’s a process to developing things that have the potential to kill their users.

                  A new smartphone can go out with bugs and get updated over time. A self-driving car absolutely cannot. I think Silicone Valley is treating this the same way they treat other new tech. Let the users do the beta testing. That’s not going to work in this case.

                  Similarly, there have been several large class action lawsuits against drug makers who have rushed and/or falsified trials for drugs that they really wanted to sell.

                • Pseudonym

                  In what ways do you think the development should be done differently than it is now?

                • tsam

                  In what ways do you think the development should be done differently than it is now?

                  Of drugs or cars?

                • DrS

                  Silicone Valley

                  This is in southern California, Silicon Valley is in northern California.

                • Pseudonym

                  Cars, not drugs.

                • weirdnoise

                  Telsas already have over-the-air software updates. Of course, this doesn’t mean that a new fog-piecing radar unit will suddenly appear on your car, but one thing some commenters seem to be ignoring is that driverless cars can have sensors that utilize wavelengths and modalities of operation than aren’t available to humans. I’m not particularly confident that driverless cars will ever happen (certainly not in the 22 years actuary tables say I’m likely to be alive), but that won’t be because technological solutions won’t ever exist.

          • tsam

            There’s a perhaps utopian dream that self driving cars/trucks will put an end to crashes. (Drunk drivers can sleep while driving, distraction and fatigue will no longer cause wrecks).

            But autonomous, self driving cars are almost certainly never going to happen, and they won’t stop human error (especially given that human error is built into the software and hardware).

            However, the systems that are coming out now (auto braking, lane drift warnings, etc) should, over time, reduce collisions. Seems like they’d be better suited to develop those and install them in every car, rather than try to jump ahead to 2050 with a technology that needs a road infrastructure to be a certain way for it to be reliable.

            • brugroffil

              IIRC there was a big announcement a week or two ago that by 2022, almost every new car sold in the US will have auto-braking.

              • tsam

                Excellent.

                (Not a Monty Burns excellent, but a Spiccoli excellent)

        • mds

          For some reason, a lot of people, including you apparently, think it is just fine that 10’s of thousands of Americans die when we could make the roads safer.

          Yeah, self-driving cars will finally lead to all the infrastructure improvements we’re in desperate need of, and which are part of the reason the roads are more dangerous than they need to be. The alternative would be some mysterious new technology that moved people in groups with a much higher rate of safety than automobiles, and with much more efficient use of energy. Maybe we could put Google on that conundrum. Their people could think about it while riding one of the company shuttle buses to work.

          (I agree that it’s disgusting that Loomis doesn’t support increased spending for infrastructure, including mass transit. It shows he has no actual solidarity with all those poor service workers who would otherwise be whisked to their jobs in their own self-driving Volvos.)

          • Pseudonym

            Because those environmentally and economically efficient Google shuttle buses are so beloved by progressives of all stripes.

            • mds

              Well, I hate them because a company involved in massive tax-avoidance and indirectly increasing lower-class workers’ commute times have created a separate private transit system for themselves that free-rides on public infrastructure, but yeah, most progressives are attacking them because they’re environmentally and economically efficient, not because they’re watching private pods full of highly-paid devs whizzing around as BART sinks into the shit. Disappointing, I must admit.

              • Pseudonym

                The free-riding is a legitimate problem, but I don’t see a lot of support for just charging a fair price for private shuttles or for developing a public transit system that would provide the same service. Lots of the opposition comes from people who just don’t want tech workers moving into the city and driving up housing prices and the cost of living. That’s an understandable point of view but not one that is capable of being addressed by any “mysterious new technology”, so I’m not sure what your point was. You don’t like Google’s self-driving cars, you don’t like Google’s private shuttle buses. Is it just because you don’t like Google because of their tax avoidance schemes? Do you think Google is a significant obstacle to building better public transit on the peninsula? Do you think highly-paid devs don’t ride BART or Caltrain? Would you support improving public transit if it meant more Silicon Valley tech workers living in SF?

              • Richard Gadsden

                My objection to Google’s buses is that they should run from San Antonio station to the Googleplex. If they put Google’s political and financial muscle behind improving Caltrain, then that could easily be a 10-12 car train every ten minutes instead of a 5-car train hourly, and it would improve service for everyone, and not just for Google.

        • njorl

          Car crashes kill about 1.3 people per 100 million miles driven. If automated cars reduce that number significantly, it will only be because people won’t tolerate them otherwise. Safety is not the primary reason people want them.

          They are desired because they could conceivably make traffic significantly more efficient. A large number of cars using a single efficient solution for getting from A to B will do it much more quickly than a large number of people all working without much knowledge of what the other drivers are doing.

          They can do things like safely tailgate at high speed, merge from 2 lanes to 1 without slowing significantly, coordinate with traffic lights so that flow through an intersection is optimized etc. Safety isn’t the goal; it’s a requirement.

        • urd

          And I take it if such cars do exist you’d be in favor for making such cars the only ones available?

          Because unless you do that, car accidents will continue to exist in large numbers even with self driving cars on the road.

          And even then I’ not so certain; see Major Kong’s point up-thread.

      • xq

        Technology already has made cars a lot safer than they once were.

        • There’s no question about that. The question is, why should we assume that this particular technology is necessarily going to make cars appreciably safer (or safer at all)? Compared to many of the technologies that have in fact made cars safer, “self-driving” appears to have many more possible failure modes that might in fact add to the danger. (To be fair, it turns out that airbags do too!)

          • Philip

            Because driving is, in many ways, exactly the kind of activity humans are bad at and computers are good at. It requires sustained attention, over long periods of nothing in particular happening. Computers don’t get sleepy or bored, or get distracted by a ringing phone, or turn around to yell at the kids in the back seat. There’s a limit to how much you can fix the problem of “humans cannot pay attention to something boring for long periods of time” without removing the human from the situation entirely.

            • Because driving is, in many ways, exactly the kind of activity humans are bad at and computers are good at

              This statement is exactly wrong. Driving is the kind of activity that humans excel at and computers are bad at–the kind requiring real-time navigation of space with a contextual understand of the surrounding environment.

              People are actually pretty good at navigating the real world. Personally, I find the whole desire for robot chariots to be somewhat pathetic.

              • Personally, I find the whole desire for robot chariots to be somewhat pathetic.

                Only because you haven’t thought through the tremendous potential for Robot Ben-Hur!!!

                • N__B

                  I look forward to robot membership in the NRA.

              • Philip

                This statement is exactly wrong. Driving is the kind of activity that humans excel at and computers are bad at–the kind requiring real-time navigation of space with a contextual understand of the surrounding environment.

                The core navigational problems are things computers are capable of doing, and need a lot of work on. The core attention problems are fundamental to human biology. Unless your proposal is cyborgs, the computers’ problems can be fixed, but the humans can’t.

              • Driving is the kind of activity that humans excel at and computers are bad at–the kind requiring real-time navigation of space with a contextual understand of the surrounding environment.

                The parts you’re missing are bordom, fatigue, and distraction as well as overconfidence and an environment with speeds and constraints were not evolved for. Those are things that are hugely difficult to improve in people. Real time navigation of space etc is something we are getting better at programming all the time. Now, it’s not clear when the cross over point will happen..

              • Thirtyish

                This statement is exactly wrong. Driving is the kind of activity that humans excel at and computers are bad at–the kind requiring real-time navigation of space with a contextual understand of the surrounding environment.

                Except…this isn’t true. Human beings are notoriously unreliable. They have short attention spans and are easily distracted both by external stimuli and their own internal thought processes. They are easily influenced by moods. They are restless and can quickly rationalize not following certain road traffic rules that have been codified for over a century just because they didn’t feel like following that rule at that particular moment. In short, humans are not all that safe for the road.

                That being said, I don’t trust self-driving cars, either. They’re programmed by humans, after all, and so much can go wrong, catastrophically. I don’t have a solution. Driving is not a safe activity is ultimately what I’m getting at, now matter how it evolves or is operationalized.

            • Murc

              It requires sustained attention, over long periods of nothing in particular happening.

              Here’s the thing; it actually doesn’t.

              Most people enter a semi-catatonic state when driving, especially a route they know well, like their daily commute. There are documented cases of people navigating perfectly through heavy traffic from one location to another while hopped up on Ambien or otherwise not wholly conscious, but even disallowing those most people are not paying attention while they drive. Their brains kick over into a state of waking sleep.

              (The same thing happens while walking or running a route you’ve done a million times; your mind sets your body on autopilot and goes elsewhere.)

              This is actually the cause of a nontrivial number of traffic accidents; it takes the brain a short but measurable time period (I think it’s like a quarter-second) to exit that state and return you to full awareness, and an awful lot can happen in a quarter-second on an icy road.

              It’s also why intersections and corners you’d think would be deathtraps aren’t; people who have to execute a complicated merge at a highway interchange they know is dangerous will “wake up” to do that, as will the other drivers around them. It’s the patch of ice they don’t know about or the guy who runs a red light they can’t react.

              • Philip

                Well, that’s what I mean. Driving safely requires constant attention. Because, for example, people often can’t see things they don’t expect unless they’re fully alert. That’s why people drive right through pedestrians, bicyclists, or even motorcyclists. They literally never see them because they’re looking for other cars.

                Plus, the self-driving car won’t be actively trying to kill me for being on a bike.

              • TrexPushups

                Yep I have ended up driving to the wrong place more than once while remembering little to nothing of the drive.

                I freaking hate driving. I would much rather be using my time to think about something else without risking death & dismemberment.

              • steverinoCT

                I have a long commute on a fairly open highway and will find myself cruising along when Bink! I pop to full alertness because the car ahead slowed to cross an invisible distance line (curiously, always almost exactly two seconds– I keep myself amused by checking these things). And other triggers like that.

                I have found I am a very good driver, sometimes. And sometimes I am a bad driver, without intending to be. I wouldn’t mind some sort of automous highway driving; a wire embedded in the road to follow rather than some visual thing, short-range radar for distance-keeping, inter-car communication linked with GPS units to get in a “train” of cars until time to make an exit and resume hands-on control.

                I make onsite visits from my office so I can’t take mass-transit; I need a car.

          • xq

            I think the idea that there’s no future version of self-driving cars that are safer than human drivers is pretty implausible. The technology will improve with time, reducing those “failure modes”; human error probably won’t decline much with time.

          • Rugosa

            Yep, I’m in the death zone for the airbag in my car (5 feet tall on a good day), so the 300-lb guy whose personal freedom requires that he not use a seatbelt can be saved from his own stupidity in the event of a crash. The last time I bought a car was 10 years ago. I looked into the issue, and there was some talk about installing an on/off switch for the airbags, but I think the nanny state* won this round.

            *actually, probably insurance companies, but let the right have a bone

            • GFW

              This is a weird idea, but do you suppose that running some sort of bungie cord across the steering wheel could slow down the expansion of the (center of the) airbag just enough to make it safe for a smaller person? The potential problem is getting hit by the added mass of the bungie itself. Ideally the explosive charge would be variable and tuned to the size of the driver by a weight sensor.

              You can get an on/off switch. I don’t know how much it costs, but here’s how: http://www.safercar.gov/Vehicle%2BShoppers/Air%2BBags/ON-OFF%2BSwitch%2BRequests

              • Rugosa

                Thanks. I’ll file that away for the next time I buy a car.

              • so-in-so

                Not to mention the rather nasty (often metal) hooks on the ends. Picturing those being extracted from my upper body is discouraging. The cut-out switch is a better idea.

                • GFW

                  Well, I was assuming the hooks could be affixed to the steering wheel in a way that they wouldn’t be popped off by the airbag – but any assumption, including that the cord wouldn’t just detach from the hook, is dangerous in the absence of testing. So yeah, the on-off switch would be much better.

              • tsam

                running some sort of bungie cord across the steering wheel could slow down the expansion of the (center of the) airbag just enough to make it safe for a smaller person?

                Please don’t do this. The best way to handle this is find a way to sit further from the center of the steering wheel. I know that’s not easy, but car selection is probably the biggest factor.

                A bungie cord is no match for an airbag, and all you’re doing with this is making two possibilities for your fingers to get snagged in the cord while you’re turning, and possibly getting whipped by the cord if one side comes loose.

          • so-in-so

            More possible failure modes than a random human driver? When we have all seen drivers totally distracted by texting, reading books or papers, applying makeup, etc.? Not to mention getting sick, sleepy, being impaired in some other way or just deciding to do something stupid?

      • NYD3030

        It’s not as if American companies are going alone on this one, every single auto company is invested heavily in self drive as far as I know.

        You could maybe make the argument that silicon valley has just snookered Daimler and Toyota. But I don’t really think so.

      • Don’t be “skeptical”, check it out. Google have run up hundreds of car-years of trials without any serious injuries. They only just had their first accident of any kind that could be attributed to the software (pulled out too slowly in front of a bus). It works and is safe. Te problem is the high cost of the radar. The software would be amortised over large numbers.
        Volvo are trying to catch up with an inferior product. Google’s maps don’t need new white lines.

        • ColBatGuano

          Google’s trials in no way represent a real life challenge.

      • So, aside from techno neatness we have:

        1) Safety (people are terrible drivers in all sorts of ways; remains to be proven in real world conditions but the possibility is clear)
        2) Reduce wasted time (lots of driving isn’t for pleasure or pleasurable so if we can reclaim that time it’s a win; downsides include increasing congestion/absurd commuting and job loss)
        3) Reduce the size of the fleet or certain externalities such as close to shops/jobs parking (car scan drop us off and pick us up making parking or car sharing more feasible).
        4) Efficiencies (such as gas use)

        Roughly, the upside is more public transit like car use. That could be a downside companies to Real public transit.

        And while the strides made are incredible (12 or so years ago at the first DARPA grad challenge, no e try finished; no we have cars on the road), there’s a pretty big hill to climb. That last 90% won’t be easy.

        So, the potential is great, but the opportunity for crap is great (sprawl out the wazoo, etc)

        Since it could prompt a faster turnover of the fleet, the car companies are going to be gung ho up to the point where it might trigger a smaller fleet or new competitiors.

        I think it will happen to some degree and relatively soon. But per usual it won’t be utopia. And there will be a lot of unforeseen aspects.

        (Servicing these things will be interesting. I Imagine we’ll see a lot of assisting feature long before any full automated cars hit the market.)

        • leftwingfox

          5) Access. The ability for disabled or impaired individuals to have access to reliable independant transportation.

          I have a number of friends who are excited about self driving cars because they have ADHD and don’t trust themselves behind the wheel of a car.

          • TrexPushups

            Ding! Ding!

            Exactly why I hate driving. I can’t stay focused my mind always drifts and I miss turns etc.

          • Chuchundra

            Or old people. There are a lot of elderly drivers who should not be behind the wheel, but who are reluctant to give up the freedom a personal vehicle provides.

            • Access for people who can’t drive themselves is, indeed, a big one.

    • jim, some guy in iowa

      people believe in safety and they believe in efficiency which they think we’ll have more of with self driving cars

      the thing pictured looks worse than useless for where I live, especially in winter but also in spring when the frost heaves the gravel roads

      • so-in-so

        The military has been applying the technology to off-road vehicles, so there is no reason to think a pickup truck or SUV can’t also have the self-driving technology.

    • Warren Terra

      Well, I don’t read the business press, but the clear darling of the business and especially technology business world in the general media for the last couple of years has been Uber – a business where users don’t have to drive themselves around, because of technology (and displaced workers). The runner up would be Tesla – fancy technological cars basically no-one owns or even sees from week to week. So while I’d never presume to have a valid opinion on what the public wants, self-driving cars are totally a sweet spot for getting enthusiastic media coverage.

      Some of this may be who writes about business and technology: a lot of them are young people in New York and Palo Alto who travel a lot and ride in a lot of cars driven by other people.

      • Steve

        Uber is investing heavily in technology for self-driving cars for obvious reasons.

      • Halloween Jack

        The runner up would be Tesla – fancy technological cars basically no-one owns or even sees from week to week.

        Elon Musk just introduced a new model that’s about half the price of the previous one, and has north of a quarter-million preorders for it, even though it won’t arrive until late next year. So, that could change.

    • malraux

      Lots of people want the advantages of taking the bus or subway, but don’t want to share space with others.

      More pragmatically, driving is one of those things that humans are awful at. it requires constant attentiveness, but in the majority of cases isn’t very interesting. That’s the sort of job that automation in theory ought to focus on.

      • so-in-so

        I’m not sure if the issue is “not sharing space with others” or if it is “I want to go when I WANT to, not on some schedule”. Cabs sorta work for that, but often are hard to find except in major cities.

        • TrexPushups

          exactly! Also nice to be picked up at your house or wherever you are.

          A fleet of self-driving cars and standardized cargo holders would let people dispense with the expense of owning a car and simply get the transportation they need.

        • ColBatGuano

          It seems like self-driving cars and an Uber-like (without the greed) app is going to be the end result. You’d type in your phone/implant your location, end point and time and various options would pop up. Choose and go.

    • Patrick

      I’m a self driving car doubter. But as someone who regularly bikes to work, I sure as hell wouldn’t mind replacing the texting / somehow drunk at 7am / thinks its funny to ride people off the road drivers with a robot. Sure, it’s only 1 in 10,000 drivers that suck that much, but that one can really really ruin your whole day.

      I’m even guessing they’d be programmed to stop and call 911 on an impact, instead of fleeing the scene and calling in their car as stolen in a few hours.

      • njorl

        “I sure as hell wouldn’t mind replacing the texting / somehow drunk at 7am / thinks its funny to ride people off the road drivers with a robot.”

        That behavior can be programmed in, I’m sure.

    • Matt McIrvin

      Fundamentally, I really fucking hate driving. Having to do it every day takes a massive psychological and physical toll. I know some people enjoy it, but nobody could enjoy it under the conditions where I usually do it.

      Mass transit is a solution for some people. It’s not a solution for all Americans unless we depopulate most of the country, which I guess might eventually be necessary anyway.

      • witlesschum

        I really fucking hate commuting or driving for errands. Taking a long car trip can be a pleasure or just driving to sightsee or explore.

      • Thirtyish

        I love living in NYC, where I don’t need to drive (I don’t even have a car here–good riddance!). But I share the strong hesitation about the idea of self-driving vehicles. I just fundamentally don’t believe that the technology could ever be foolproof enough to trust. Not that I particularly trust human drivers, either. Driving in general is tremendously dangerous no matter how you slice it.

        • Steve

          Never? I think that is to strong.

          The interesting point will be if/when self-driving cars are on the road with human drivers…humans don’t necessarily know how to respond to the movements of self-driving cars (assuming they have a logic all to themselves). I can imagine, at least in the interim period, a large spike in accidents arising from this.

      • Steve

        I suppose self-driving cars allow for the continuation and spread of the suburbs (especially if they can coordinate in such a way as to reduce traffic congestion). So they might actually produce more sprawl on net even if they cut down on the amount of space given over to parking lots, parking spaces, parking garages, etc.

    • JL

      I would love to have a self-driving car for myself – note that I said nothing about public policy or technological feasibility here – because I hate driving to the point that I’ve never gotten a license despite having had a permit three or four times, but I suspect I’m not representative of the general population.

      As other commenters point out, the bus and the subway exist (in some parts of the country, anyway), as do bikes, walking, and the commuter rail, and I currently get around overwhelmingly on the combination of those things, with Lyft, and, when applicable, carpooling with others, to make up the difference, because sometimes public transit does not actually go where you need to go (or the combination of routes is such that it takes two hours to make a trip that would be under half an hour in a car, or it’s the middle of the night and public transit stopped running an hour ago).

    • addicted44

      Most people think of self driving cars as basically cars without drivers. They are that, but they are a lot more.

      For one, self driving cars will be the only form of mass transit Americans invest in. Yup, mass transit. When (not if, unless holograms beat them out) self driving cars happen, private car ownership (except for the very elite) will end. No longer will people spend tens of thousands of dollars upfront, and hundreds of dollars on fuel and maintenance monthly, to buy something whose utilization is below 10%. Self driving cars will be provided as a service where you input your ride, number of passengers, and baggage, and the appropriate car will show up at the right time.

      Self driving cars will greatly reduce traffic deaths. Which, btw, far exceed deaths due to gun violence in the US and the rest of the world.

      Self driving cars will be far more environmentally friendly. They will almost certainly be electric, will almost certainly be driven by renewable energy (it would be cheaper for a provider to setup their own charging stations than to get electricity off the grid) and would be more efficient (you won’t usually see SUVs transporting a mother and her child). Safer driving will mean that a lot of the safety features which add weight and dramatically reduce mileage won’t be necessary anymore, further improving efficiency.

      Liberals can either accept the almost inevitability of self driving cars (one thing j guess Atrios and Erik don’t seem to realize that we ALREADY have self driving cars…the Google cars, obviously, but even the Model S can drive itself…Tesla added this feature with a software update) and work to make sure the benefits are evenly distributed and they are regulated in a fashion that causes minimal disruption.

      Instead, apparently the left wants to take the Luddite approach, and then whine when the assholes drive their adoption leading to self driving cars becoming another tool in crushing the non-elite class.

      • Hey, another commenter who doesn’t understand what Luddism was.

        • Murc

          Words change in their meaning. A44 is using Luddite in a perfectly acceptable modern sense.

          • Words indeed change their meaning. But not proper nouns.

            • Murc

              Interesting. So your actual issue is with “Luddite” as opposed to “luddite.”

              Makes sense.

              ETA: I actually do mean that it makes sense. That looked a lot more bitchy than I thought it did when I hit post.

            • So Shakespeare had it all wrong when he has Hamlet say “It out-herod’s Herod”?

              Of course proper nouns change their meaning (“Communism”) and acquire mew connotations ans wel as adjectival and verbal forms. Erik’s reactionary linguistic essentialism places him clearly behind the wheel as a language Luddite.

        • Ahenobarbus

          As Dilan Esper correctly points out elsewhere in comments,

          Luddism is no longer a literal term. Words aren’t static…

          When someone says “Luddite” today they aren’t actually engaging in a discussion about early 19th century English textile workers.

          • Except that Luddite is a thing. You can’t just invoke proper nouns that are real things for whatever you want. Language changes all the time, but this is an area where using the term just makes you look stupid because, no, Luddism is this. If you want to turn it into an adjective, I guess it’s slightly less bad. But those who use it at all can’t really define what they are talking about. It’s just a pejorative in this sense.

            • Ahenobarbus

              So what does “Futurism” mean? You know that was really a thing. You’ve used it a few times in these comments as a simple pejorative.

              • I was making no claims in those comments that had anything to do with the Italian artistic movement.

                • And you also only used “futurist” in lower case; a use that’s been common since (well before) Alvin fuckin’ Tofler, for god’s sake. Your interlocutors are (surprise) not overflowing with good faith, I think.

            • tsam

              I just spell it with a little l. Problem SOLVED, M8!

              No need to Whig out about it.

              • Henry Clay always gets the blood pumping

              • FlipYrWhig

                You talkin’ to me?

        • Pseudonym

          Would you consider nitpicking word usage a meaningful response to any of your comments?

      • Matt McIrvin

        Erik is “the left” now?

        I personally think it’ll be a long, long time before self-driving cars can deal with, say, Massachusetts (he said, looking out the window at an April snowstorm). In general, the last 0.1% of situations are going to be tougher to deal with than the other 99.9%, and 99.9% isn’t going to be good enough for everyday use as anything other than a gimmick for specialized situations, like the Tesla autopilot.

        I think they would be desirable things on balance, though there would be social disruptions. I unfortunately suspect the displacement of truck drivers that Erik is worrying about is going to happen long before full-time self-driving cars become a reality for the masses, because making a truck that can, say, shuttle between specially instrumented transfer stations on an interstate is an easier problem, and that removes the need to pay drivers for that stretch of highway; they’d just do the hard bits.

      • urd

        Here we go again with the Luddite bullshit.

        I noticed this is a favorite term of the tech “disruptors” like Uber and their ilk (and their followers) when anyone challenges their view of how things should be.

        By the time self driving cars are a reality, many of the cities now bending over backwards to accommodate them will be underwater.

        Deal.

      • Pseudonym

        1. As mass transit: this has some validity. It’s not as efficient as most other forms in many ways, but it does use the pre-existing road network. Fuel and maintenance costs are generally proportional to miles driven and so won’t necessarily change with increased utilization.

        2. Reducing traffic deaths: the technology developed in pursuit of self-driving cars is going to be used for safety-focused driver augmentation long before the cars become truly autonomous at scale, so this isn’t an argument in favor of driverless cars per se.

        3. Environmental friendliness: this is really a subset of your first point. Self-driving cars aren’t particularly more likely to be electric or fuel-efficient or space-efficient if the usage and ownership pattern doesn’t change to be closer to the lines of existing mass transit. It’s unlikely that physical safety standards will be lowered, if for no other reason than that there will still be driver-directed cars on the road too for a very long time.

        Self-driving also doesn’t have to mean driverless any more than planes with autopilot or UAVs are pilotless.

    • Murc

      This might seem like a dumb question, but why do so many people think we need self-driving cars?

      We don’t need self-driving cars anymore than we needed regular cars.

      But they’d be amazing to have. I’d fucking love to be able to summon a robot car, take it to where I want to go, and then when I need to leave summon another robot car to take me away. And while I’m traveling in them, I don’t have to drive, but can do other things.

      This sounds amazing to me.

      Fuck regular cars. Bring on our robot overlords. The faster, the better.

      • I don’t have to drive, but can do other things.

        Like whatever your employer tells you to do since that is who will be taking over that time.

        • Murc

          This response seems like a non-sequitur. If they try to do that, why not fight that battle then, rather than say “gaining more time and less stress for yourself is pointless since it’ll just be reclaimed.”

          Indeed, we shouldn’t even wait until then; the fight to prevent your employer from claiming uncompensated time from you is here, now, and ongoing.

          • urd

            Not really.

            How well has that battle gone so far? Most firms in the US have steadily chipped away at personal time with the assist of technology. I think it naive to believe that suddenly with this technology workers will be able to hold the line. It could make things even worse: I can see this as a way to cut back on telecommuting as getting to work is no longer an inconvenience.

            In any case, for the vast majority of people the fight has already been lost. Hard to keep fighting a battle that is already over.

        • MDrew

          WTF? So what does this have to do with SDCs? I guess we might as well just eschew any technology or hell personal organization steps that will save us some time in our personal lives, because it’s a futile endeavor since our employers will capture every minute we net?

          Because that’s the only logic of this argument.

          • It’s not an argument against the cars. It’s argument against saying that the cars will free us.

            • MDrew

              Actually as was my point it’s even less specific to the cars than that. iI’s an argument against thinking that there’s any way to better organize our lives, certainly through technology, to free up time to do things we want, because employers will claim every new free minute. Which is the false and ridiculous argument of a flailing ideologue.

              • I’m not making any such argument. I am saying that like with e-mail and smart phones, this is in fact what is going to happen.

                • MDrew

                  You’re making any such argument, you’re just saying that what I say you’re saying will happen will happen?

                • ColBatGuano

                  It’s not an argument for Eric. It’s a truism that will happen no matter what.

      • Wapiti

        Frankly, that sounds like Uber, or a responsive taxi system. Does it really matter if the car is robotic or driven by a competent human?

        • Murc

          I trust the robots more than I trust other humans, and I don’t want to share the car with some dude I don’t know. I’m fine with that on subways and buses, and I’m okay with it on the rare occasion I take a taxi, but if I’m summoning a robot car for normal everyday travel I want to be able to do things like rock out to loud music or yell at the radio or scratch myself in socially inappropriate places. The stuff I do in my car now.

          • But what if you want to scratch some dude you don’t know in socially inappropriate places???

            • leftwingfox

              Then robot drivers don’t judge.

        • Steve

          Will some subset of the cars be programmed to be racist assholes who speed off when they see you?

          • tsam

            Heh–lock the doors when they see a black person? That’s not funny, but it made me laugh. Going to flagellate now, brb.

      • Brett

        I’d love to get one with a more spacious interior and a “lounge” style setup with a table in the middle. You’d still need everyone to wear seatbelts while it’s driving, but on very long drives you could play board games, eat at the table, etc.

        • tsam

          Also sexy time with the right group of oh shit I’ve said too much.

    • Karen24

      Well, my 79-year-old widowed mother would very much like to be able to drive to Austin and visit her grandsons, but she can’t drive that far alone. A self-driving car would be a godsend for her and the rest of our aging population.

      • Jordan

        Right, that is another good benefit.

        • FlipYrWhig

          It capitalizes on old people’s high levels of trust in the benefits of new technology.

    • Brett

      Fully automated cars could theoretically not only be much safer (most accidents are caused by driver error), but run at higher speeds and be extremely useful for people who are impaired in terms of mobility. I know my grandmother would have wanted a driverless car once she could no longer safely drive, for example – she was totally dependent on people bringing her stuff and hated it.

      If one was available and in my price range, I’d buy one in a heartbeat. The vast, vast majority of my trips in cars are of the regular variety, like driving to work and the grocery. I’d be nice if I could just get in the car in the morning, press “work” on the menu, and let the car drive me there while I ate a quick breakfast.

    • Richard Gadsden

      Über wants to stop paying drivers.

  • marduk

    Always good to go on the record with a prediction that’s guaranteed to be proven wrong within a few years.

    • jim, some guy in iowa

      what prediction is that? the one that says the states will start spending money to paint lane markings when they won’t spend on paving roads? Christ, here the fucking koch bros are financing primaries against republicans who voted to increase our fuel tax just to (almost) stay where we are let alone improve the roads

      • addicted44

        The fact that Volvo’s cars require lane markings to drive themselves is only evidence of the fact that Volvo isn’t gonna be the ones making the future self driving cars. This is like saying the Blackberry Storm sucked therefore Apple’s iPhone will also be a failure.

        There are many approaches real companies that are doing the research here are taking and very few involve lane markings. Google, for example, uses its mapping technology to drive it’s cars (you don’t think the goal of Street view was simply pretty pictures on their maps, do you?). Uber has pretty much taken over the entire CMU robotics dept in an effort tondevelop their self driven car.

        If humans are able to drive without lane markings, why do you think computers cannot do the same?

        • jim, some guy in iowa

          my point was about infrastructure investment by the states- the computer will *have* to be able to get along without that

      • marduk

        The lane markings are a red herring- a unique failure of Volvo’s design.

  • NonyNony

    “Ain’t going to happen” meaning “never”? You’re probably wrong on that.

    “Ain’t going to happen” meaning “not in the next 40-50 years”? You’re probably correct on that.

    The technology is not nearly as good as the people touting it have been claiming. It relies on comprehensive mapping of the areas that stay up to date and are marked with construction outages and sensor networks that require well maintained, well lit, well marked roads. And that’s before we get into the political changes required on a state-by-state level to allow for the cars to be used.

    The technology will improve and eventually it will be able to do the things that its needed to do on our poorly maintained and poorly designed roadways. But it won’t be in the next 10 years, and it likely won’t be in the next 20 either. (And then the political hurdles will need to be tackled – it’ll be fun to see the automakers struggle with 50 different state legislatures to get their cars approved for use in those states).

    • Philip

      I’d say 20-30 years. Computer vision has come a long way.

    • addicted44

      The mapping approach is Google’s approach (and likely not their only one). There are other approaches which may be more successful quicker.

      • When every car is a mapper and uploads data constantly on the condition of the roads, congestion, and traffic hazards, the map will be pretty accurate.

        • Chuchundra

          It’s already pretty accurate if you use an app like Waze. If every car was reporting real-time road conditions back to a central server, it would be nearly perfect.

    • Gregor Sansa

      I’d guess, 10-15 years until it starts to take off, 15-20 until it’s everywhere. Far longer than the boosters think, but far shorter than “40-50 years”. I have an above-average understanding of the state of the art in machine learning (PhD student in statistics) and I correctly called the current slowdown in Moore’s law as early as 2008.

      • NonyNony

        That’s fine and all, but I’ve been keeping up with the peer-reviewed research (and have a PhD of my own if you’d like to whip yours out to compare sizes). Absent a groundshaking revolution in the field I think 20 years is WAAAAY too generous for being used “everywhere”. The idea that these systems will be so affordable that every Ford Focus rolling off the assembly line is going to have one installed in 20 years is already a non-starter.

        And on top of that – it’s going to take at least 20 years to get the politics of it all worked out after the tech works.

        • Gregor Sansa

          My comment wasn’t intended to say “bow down before my knowledge” but just “I’m not entirely ig’nant”. (I don’t have a PhD yet.)

          The technology doesn’t have to be cheap enough to put in a Ford Focus. It’s good enough if one self-driving Ford Focus costs less than one on the road and one in a parking space (including the cost of the parking space).

      • jmauro

        It’s probably closer to AI (since really it’s an AI program) where it’ll always be just 10 years away.

    • so-in-so

      Has there been any other major technology innovation that reached this stage (several independent working prototypes) that took 40-50 more years to reach production? Actual question, I can’t think of one off hand, but there may well be several.

  • BGinCHI

    I’ve been scratching my head about this for a while. I thought it would be a passing thing, but the declarations that driving is all going to be automated in 10 or 20 years strikes me as insane.

    1. Have any of these people been to, say, Buffalo? You can barely drive there as a skilled operator, much less as a machine. They would have to completely reconstruct the city in terms of planning and infrastructure.

    2. People fucking love cars and driving. Jesus, are there people in suits who don’t know this? Would some dudes (or ladies) who watch NASCAR, or restore cars, or are just into what they drive going to leap at the chance to ride around hands-free in a 600 lb flashlight with wheels?

    Hells no.

    The market for this probably is no larger than the Bay Area, and might not even work there.

    • NYD3030

      The machines will eventually be superior to the most skilled operator, and while people do love car culture the safety, convenience and decreased cost of autonomous vehicles will be pretty hard to resist no matter how much you grieve for the loss of Dale Earnhardt.

      • I’m curious as to what you think all the laid off truckers should go do for a living.

        • Dilan Esper

          What did all the laid off buggy whip makers and carriage drivers do for a living?

          You are dangerously close to Luddism here.

          • First, you don’t even know what Luddism is.

            Second, in a society where the last decent paying job for people without college educations is trucking, what will replace it?

            Rather than cheap snark, answer the question.

            • Dilan Esper

              Luddism is no longer a literal term. Words aren’t static any more than economies are.

              I dispute the premise. Trucking is the only good job?

              And people have bemoaned job loss for decades, but we seem to keep getting richer. At any rate, that’s not a good reason to kill tens of thousands of people.

              • Please list the categories of good-paying jobs for people without college educations.

                Also, accusing someone of something that has an actual meaning means that you have to understand what it means. This is like accusing Barack Obama of Marxism. No, Marxism is an actual ideology and system that shares nothing with Obama. Luddism was a social movement with actual people who believed things.

              • Warren Terra

                Luddism is no longer a literal term. Words aren’t static any more than economies are.

                Yeah, screw this notion. There are perfectly good words to describe technophobes and the hidebound. Defaming the good (if likely apocryphal) name of Ned Ludd and ignoring the valid and indeed highly resonant compaints of the displaced workers in his movement (who were opposed to being turfed out with nothing, not to technology itself) just makes you look like an ignoramus.

                ETA:

                people have bemoaned job loss for decades, but we seem to keep getting richer.

                Who’s “we”? It’s not the people in the vast majority of the income distribution in this country, not for decades.

                • Dilan Esper

                  Actually working class people are a lot richer now than in the past. Material comforts being cheaper counts.

                • Brett

                  Or rather, their problems aren’t being driven by technology and automation. What’s hammering the poor is housing, transportation, and health care – the former comes from bad housing policy, the second from bad transportation funding policy, and the third from bad government health care policy.

                • Pseudonym

                  And education too, but thankfully Sebastian Thrun is fixing that now that self-driving cars have been solved.

              • rea

                If unemployed truck drivers were to smash self-driving trucks, that would be Luddism

                • And actually, I don’t really see why they shouldn’t do exactly that if push comes to shove. I mean, it won’t work, but still.

                • Brett

                  It won’t work because the shift isn’t going to be overnight. They’ll have drivers on for liability reasons first, then have ride-along staff for security reasons (a fully automated truck is going to be a tempting target for burglary unless you build it like a lockbox with cameras).

                  By the time nobody is riding on these things, it will have been years since they had regular drivers who might be displaced if it happened over night.

            • addicted44

              That’s a good question. Driving provides a lot of good jobs. And it’s important how we figure out what we are gonna do about work and distribution of wealth in an increasingly automated work.

              However, the loss of jobs isn’t gonna stop the actual technological progress, nor the implementation of that technology.

              Frankly, the automation trend is only going to accelerate, and it’s about time we start talking about an economy which shares the benefits of automation with everyone and not just the capital holders, and start talking about what the fuck are humans actually supposed to do with their lives when we are no longer needed to spend 50+% of our waking lives doing jobs.

            • leftwingfox

              Guaranteed Minimum Income.

              Yeah, I know, “Morally correct but politically impossible”, but how long are we going to accept that as anything but defeatism?

              • Ormond

                Right? Maybe we should stop fighting the last war that we already lost and start engaging with the social problems as they are and deal with automation. The train left the station, lets organize as to where it’s going.

              • ColBatGuano

                “Morally correct but politically impossible”

                Well, politically impossible up to the point that 50.1% of the population’s kids are starving and the government’s response is Ted Cruz theocratic homilies about family.

            • Pseudonym

              Luddism was a movement of skilled workers who protested their displacement by businesses who used new labor-saving technologies to replace them with machines and less-skilled and less-well-compensated workers, right? How does this not apply? Would Neo-Luddism be a more accurate characterization? Does complaining about the usage of the word “Luddite” actually solve technological unemployment or discredit the notion of the “Luddite fallacy” by whatever name it’s known?

        • NYD3030

          I think they’re fucked, frankly, and I have no idea what a good solution would be. Basic guaranteed income sounds like maybe a good idea but I’m also skeptical of it on the grounds that work is an important part of human life independent of its economic necessity and also any program that gives people ‘money for doing nothing’ seems pretty easy to kill.

          So nothing but bad news on that front. I think autonomous vehicles will be a reality rather quickly, at least for interstate travel. And I have no idea what to do for the people whose lives are destroyed by loss of work. But I see no historical or political indication that anything will be in done to preserve these jobs.

          • OK–but I would argue that as a strong supporter of these vehicles, you have a moral obligation to figure it out. Otherwise, it’s just another example of people demanding the poor suffer for their whims.

            • Dilan Esper

              And as an opponent, you would be morally responsible for the deaths that result from not doing this.

              • I’m sure when those truckers are committing suicide from not being able to feed their families, they will thank you for your concern about their lives.

                • Dilan Esper

                  You really believe that’s going to happen at the same rate as highway deaths?

            • Philip

              Otherwise, it’s just another example of people demanding the poor suffer for their whims.

              Good to know that “prevent 30,000 deaths a year” is a whim. As such a strong supporter of human-driven death machines, you have a moral obligation to figure out how to prevent those.

              Yes, in a perfect world we’d all be using mass transit which would in fact prevent most of those. But we’ll have fully autonomous cars long before Americans give up their suburbs and take the bus to work every day.

              • The answer to your challenge is right there in your second paragraph.

                • Rob in CT

                  I don’t see why both more density and more automation can’t happen…

                • Philip

                  And also in my second paragraph is the problem with it. Many Americans, as you yourself love saying, will never give up their suburbs, or take the bus. We have to deal with the reality that mass transit is always going to be limited here. I think there are serious obstacles to good self-driving cars. But I think that, the US being what it is, we’re vastly more likely to overcome them because there will actually be meaningful investment.

              • Ormond

                Also: allow handicapped people mobility and autonomy. Also: allow people to get to work without having to own and maintain a car.

                • Moondog

                  And the elderly.

                  Enabling people to get around also enables them to work if they need to.

                  It’s easy to imagine a lot of software-driven carpooling in self-driving taxis — as close as many cities might get to mass transit.

            • NYD3030

              I guess I’m not coming across like I intend, I am a strong believer that this will happen. And in a very narrow sense it will be good for me because I live eight hours from my family and have a young daughter, being able to visit them without having to actually do the driving part or buying three plane tickets would be great.

              I’m not sure it’s a net gain for society. There will be fewer accidents, traveling will be more pleasurable and presumably efficient. Consumer goods might cost a little less? On the other hand putting a ton of people out of work is bad for everyone and also for me, even though I don’t work in transportation. Because having gainfully employed, productive citizens who have a stake in the success of our society benifits me a hell of a lot.

              So yeah. Not so much I think it’s a glorious thing, just that I think it’s going to happen. Frankly it’s going to happen to a ton of white collar jobs as well, including most likely my own.

            • liberalrob

              …as a strong supporter of these vehicles, you have a moral obligation to figure it out.

              We have a moral obligation as a society to figure it out; but we cannot make technological progress contingent on figuring out ahead of time what all the possible impacts will be and having in place plans to deal with those impacts. That is a conservative approach to technological innovation, not a progressive one. It is putting limits and qualifications on which innovations are to be permitted and which are not.

            • Pseudonym

              Assuming that the technology is eventually developed to make self-driving long-haul trucks superior in cost and safety to human-operated ones, what is going to make paying truck drivers a more useful or politically feasible task than paying anyone else to do something useless?

              It’s not even like supermarket checkout workers; buying a self-driving car for personal use isn’t putting anyone out of a job, much less someone you have to see face-to-face. Are you going to stop selling your books on Amazon (or putting your wishlists on there) because their automation is eliminating decent-paying jobs? How about relying on research librarians instead of Google, or hand-tuned paper maps instead of a GPS? Do you buy a car without an ECU so you can pay a skilled mechanic to adjust your timing points periodically? Where do you draw the line between acceptable and unacceptable automation? Or is this like racism in school selection, where the discussion goes from blanket condemnation at the general level to “it’s complicated” at the individual level without much in the way of proposals for resolving that tension?

            • Sebastian_h

              That is a fascinating argument. What do you think of the argument that black parents should keep their kids in crappy schools instead of sending them to a nearby charter schools that they think are better?

        • Brett

          Same thing that happened to all the milk men, ice men, etc who found themselves redundant in the Postwar Period. They’ll either get sucked up into new jobs if the economy is good, or go on unemployment if it’s not (like tons of other people).

          If anything, it’ll be easier with drivers because they’re not geographically concentrated in an area where them being unemployed blows out the entire local economy (like with factory and mining towns). And they’re already used to mobility and being all over the place.

        • Barry_D

          I am curious as to why you think that your commenters won’t notice that you’ve changed your argument.

          • I’m curious why you actually think I care enough to try and fool someone.

            • Barry_D

              Whatever.

    • Scott P.

      It doesn’t have to be a big market at first. The first microwave oven was 6 feet tall, 750 pounds, and cost $5000 in 1955 dollars. But some people bought them, and that drove innovation and lowered the size & price.

      • The Dark God of Time

        The problem here is developing the software, not the sensors, cameras, GPS, and other hardware that can be improved upon by using cheaper materials or automated fabrication methods.

    • Ahenobarbus

      People love driving….until they find out (a) they’ll have to pay five times as much (or whatever) insurance premiums to self-drive and (b) they’ll be able to nap, read, work, watch TV, etc. in the back while the computer drives for them.

      I do agree that the timeline for this stuff is over-promised.

      • addicted44

        Does anyone actually know anyone who loves driving or are we just basing this on car commercials?

        Is there anyone who says, I’m gonna take this other job because it means I will get to drive 30 more minutes a day since I love driving?

        The people who love driving will still get their jollies on a race track. No one is gonna miss their daily commute, or the long ride to grandmas with the kids screaming in the back.

        • Philip

          I love driving. But yeah, no one who loves driving enjoys commuting. It’s the worst, most boring kind of driving imaginable.

          • tsam

            Yeah, this. When I say I love my car, I don’t mean I love it and driving so much. But I do like walking to the driveway and going somewhere with just my family or nobody else with me, the roads I choose, on my own schedule, etc.

        • Is there anyone who says, I’m gonna take this other job because it means I will get to drive 30 more minutes a day since I love driving?

          For about 30 years, my weekly commute (during the 9 month academic year; carpooling about half the time) was about 450 miles. Much of the time the drive (particularly, but not exclusively, the drive home…) was the best part of a workday/workweek. I actually do “love driving” in a mild, companionate-marriage kind of way. (But I hate car commercials. And I didn’t choose the jobs that required the drive: I could never find an alternative job satisfying my other requirements and desires.)

    • NonyNony

      your first argument is a good one, but your second isn’t. Because there is a tiny minority of people who love cars and driving and a larger group – especially among younger people – who find driving to be a chore, cars to be as boring as refrigerators, and really would like to have more time to catch up on Netflix or social media or whatever than drive a car.

      That’s what makes the self-driving cars so appealing to a certain crowd – the idea that you can just let something else take the tedious boredom of driving off your hands and you can sit back and either get some work done or binge watch Season 2 of Daredevil while you take that trip.

    • Brett

      2. Do you love driving all the time? I love driving up one of several canyons in the Wasatch Mountains here, listening to music. I don’t love driving to work every week day, or driving to the grocery store, or driving to various restaurants, etc. I’d be more than happy to turn over that driving to an auto-drive system.

    • N__B

      1. Have any of these people been to, say, Buffalo? You can barely drive there as a skilled operator, much less as a machine. They would have to completely reconstruct the city in terms of planning and infrastructure.

      Our robot overlords will leave Buffalo to us meat-sacks.

    • gmoot

      This. It’ll just take one self-driving vehicle that “learns” to avoid the big-ass pothole on Kenmore Avenue by driving Grandma into oncoming traffic.

      One-lane bridges are also fairly common around these parts, with social norms about when to wait and when to go. Or, how about a 4-way stop, where safe driving involves making eye contact.

      There’s also a substantial portion of the US, geographically speaking, where there are no lane lines on roads, no Google street view, no wifi/cell service, no accurate GPS mapping, etc. People do live (and drive) there.

      So, yeah, 10-20 years seems … overly optimistic. Or pessimistic, depending on your PoV.

  • malraux

    I suspect that long distance trucking will be one of the first to get some form of autonomous driving. Interstates are relatively predictable, have much more standardized markings, simplify issues relating to pedestrians or nonstandard vehicles. In contrast, city streets and personal driving involve lots of unusual driving situations.

    • Scott P.

      You could have a marshalling station outside major cities. The trucks would drive themselves from station to station, and there would be a staff of human drivers to take them the last 10-15 miles to their final destination.

      • witlesschum

        Sounds like reinventing the railroad to me. It may be my occasional driving among the semis on I-94 talking, but segregating more cargo onto rails seems like something I can get behind.

        • anapestic

          And if we had a good passenger rail system in this country, people could do all the things they want to do in their self-driving cars.

        • addicted44

          It’s exactly that.

          Self driving cars are indistinguishable from mass transit. The only difference is that they are more personalized (they get you to your door instead of the corner bus stop), and they rely on infrastructure countries have spent decades and trillions of dollars building out.

          They’re mass transit for Americas.

          • sonamib

            They’re also less space- and fuel-efficient. They require more infrastructure. It may already be built out, but the infrastructure maintenance cost will still be higher than for regular mass transit.

    • Karen24

      Exactly. Also, machines don’t need sleep and therefore can drive at 2 am, reducing traffic during rush hour and allowing ever narrower delivery schedules.

      • so-in-so

        Except, because they don’t need sleep they will be driving during rush hour AND at 2 AM!

    • Brett

      I doubt it, at least in the US. Rigs rack up a ton of miles before they get replaced, and I doubt the value of automation will be so great that they’ll simply be replaced before the existing set of trucks are worn out.

      Then there’s liability issues (as long as there’s a driver who can theoretically take control, you can blame accidents on them and not the technology), security issues with the trucks, etc. A fully automated truck is probably going to have to be a lockbox with cameras, and that might not be cheaper than simply having a ride-along guy who is there for security reasons.

      Hell, if the ride-along guy doesn’t actually have to drive, then it gets even easier. The trucks go along a particular route, you pay the ride-along guys to basically hang out in the cab (or something else) to keep an eye on the cargo (or help unload if necessary), and it all works out.

      • Increasingly it will be easy to retrofit auto driving units onto vehicles. Any vehicle that already has radar collision avoidance, power steering, power brakes, GPS and an automatic transmission has 90% of the required hardware. Add in a few cameras, and brain that doesn’t have to be any smarter than a current generation iPhone, and you’re all set from a hardware perspective. And the programming while not trivial, is a problem that only needs to be solved once (or at least once per software vendor), and then rolled out to millions of vehicles.

        • liberalrob

          I look forward to my self-driving 1929 Ford Model A.

  • NYD3030

    The self driven car required a stupendous amount of infrastructure to become a staple of American life. I’m pretty confident the effort will be made for autonomous vehicles as well.

    The alternative is staying indefinitely stuck in the late twentieth century. It’s hard to imagine given the current political impossibility of doing anything remotely productive at a national scale, but I’m pretty confident that the lure of laying off a couple million truck drivers will prove overpowering.

    If not, eventually the software will get good enough that it won’t matter.

    • The alternative is staying indefinitely stuck in the late twentieth century.

      I love the futurism that is inherent in this sentence.

      • Ahenobarbus

        I think it’s fair to ask whether you would have opposed automobiles when they first came out. Think of all those farriers and whatnots who lost their jobs.

        • This is silly–the transition of job from horse cart driver to truck driver was fairly seamless.

          Meanwhile self-driving trucks create no jobs.

          • Meanwhile self-driving trucks create no jobs.

            Hey! They create enormous opportunities for hijacking, highway robbery, and general hijinx!! Imagine a world in which the truck itself is responsible for dropping things off the back of the truck!!!

            • NonyNony

              Imagine a world in which the truck itself is responsible for dropping things off the back of the truck!!!

              Here’s my other rant :)

              Even if we get to the point of self-driving cars, self-delivering vehicles are longer off. There is going to need to be a person there to dissuade shenanigans.

              I suspect that even if we had self-driving cars right now there wouldn’t be a significant job loss because each truck would have a security guard riding in it to make sure that nobody jacked the truck. Stealing a truck when there’s a person in the mix is a risky proposition – you have to be prepared to injure or kill a person. Stealing a truck when it’s just a robot? Eh – I imagine you’d find a lot more takers willing to take that risk.

              (This is also why Amazon’s delivery drones will be hilarious when/if they really start delivering things en masse out here in flyover country. Few people are going to take potshots at the mailman or the UPS truck to nab your Amazon package, but I suspect that there will be more willing to throw a rock/shoot down an Amazon drone.)

              • Karen24

                See oceanic shipping for a parallel. Ships don’t need crews at all once out of harbor, but they still have a few to discourage people fro stealing the cargo.

          • Philip

            Meanwhile self-driving trucks create no jobs.

            Technical support, maintenance (it wouldn’t surprise me if a self-driving car has to be kept in much better repair), mapping updates, field techs (if it breaks down somewhere, you need someone to go do even the basic stuff a truck driver could’ve done themselves)… Sure, they come nowhere near the same number of jobs, but your habit of stating easily-disproven absolutes does you no rhetorical favors.

            • That’s not a job creation technology. At most, it slightly lowers the number of lost jobs and moves them significantly up the socio-economic scale. It’s a job destroyer, not a job creator.

            • NYD3030

              If you couple self driving and electric engines you get a vehicle that gets in fewer accidents and requires vastly less maintenance, not more. Electric vehicles have way way way fewer parts to break.

              • Philip

                I’m not convinced we have the battery technology for cheap electric cars, or will for decades. Current battery technologies have by-and-large hit the limits of chemistry and physics, and so far no one’s had much luck finding a new kind of battery.

            • Brett

              How do you know they won’t create similar numbers of jobs? It’s not like it would have been obvious to someone noticing that carpenters, wheelwrights, blacksmiths, and horse-trainers would be unemployed in an age of automobiles that said automobiles would produce vastly greater numbers of jobs. People were afraid of automation and machines back then too.

          • addicted44

            The whole point of innovation for the past century has been to reduce the reliance on people to do things.

            Why do you think the fact that it reduces jobs is an impediment in autonomous cars happening as opposed to being a driver?

          • …except for the long awaited rebirth of the full service gas station.

          • Pseudonym

            What percentage of Americans in the 19th century owned and rode their own horses?

          • Barry_D

            Erik: “This is silly–the transition of job from horse cart driver to truck driver was fairly seamless”

            I’m calling nymjack – this is far too stupid for Erik.

      • NYD3030

        Futurism in what sense? I’m only really familiar with the term as far as the fascist pre first world war arts movement in continental Europe.

  • Every article I’ve read about “imagine what self-driving cars could do” seems to be describing behavior designed for the express purpose of maximizing traffic congestion.

  • Denverite

    Erik, do you think we’ll see self-driving cars used in any appreciable numbers anywhere on a city-wide basis in the next several decades? It seems like insofar as infrastructure is the problem, that might be more feasible.

    • Outside of the Bay Area maybe? I’m skeptical.

      • urd

        Even in the Bay Area I’m skeptical. Outside of the areas Google and the other tech companies are really focusing on (Mountain View, Palo Alto, Sunnyvale, San Mateo, etc.) the infrastructure is a mess. Just as an example, parts of the 101 are in exceedingly bad shape after the rains we had over the winter. I hear there are similar issues in many places throughout the Bay Area.

        And some parts of the cities listed above aren’t too great either…

  • Bootsie

    When will technology reach the point where the self-driving cars won’t look like shit? That’s my question.

    • Philip

      When anyone is trying to sell them rather than make a cheap-to-produce test platform with no intention of selling it, probably.

    • so-in-so

      Seriously, does it the test unit show look that different from the Smart car or even the new VW bug? Any world in which the Cube is a viable automotive product has no need of this worry.

  • Marc

    The short answer is that self-driving cars could free up a lot of time that people currently spend driving and let them do other things, they could also dramatically lower the climate costs for transport, and they could free up a lot of resources currently used for cars. It’s also true that they could save lives.

    Many of these things do rely on people sharing, and not owning, automated cars – which can then serve the needs of many more people without needing to have one parked in every garage and at every workplace. But many of the gains don’t.

    Basically, I expect that they’ll come into use first in dedicated long haul routes and that they’ll come into use last in dense urban areas. But the logic behind them is pretty compelling.

    • Warren Terra

      The short answer is that self-driving cars could free up a lot of time that people currently spend driving and let them do other things

      This is true, but is it important? Is there any real market value for the time most people burn commuting? Not from their current employer, who is perfectly happy if they just stay at work longer or do more work at home, and doesn’t greatly care how they fit it all in.

      they could also dramatically lower the climate costs for transport, and they could free up a lot of resources currently used for cars.

      This is all to do with your point about sharing. The question is whether we’re willing to share like this (and also whether automation is even critical to this idea – see Uber). I’m skeptical, especially since the critical period is likely to be standard commuting times, when everyone wants a car at once, so maybe cars-on-demand couldn’t get all that many cars off the road unless people are sharing; mind you, systems like Uber in theory could work really well for ride sharing, but they’re already doing that, without (yet) going driverless.

      It’s also true that they could save lives.

      Again, this may be true. It’d be nice. But people are manifestly less than worried with the current rate of serious traffic accidents.

  • Rob in CT

    I’m sold on the idea that what we’ll actually see is continued feature creep. Lane departure warnings, self-stopping, etc. Bit by bit the cars will get more sophisticated at preventing human error.

    There’s a lot of upside here. A lot of people are injured or killed b/c of driving. Are there other things worth pursuing too? Absolutely (more mass transit/denser living, more working remotely, etc).

    I think a fully autonomous car would be cool, but would require that the cars, roads & signals all be networked so they can “talk” to each other. And somehow have that be super reliable and un-hackable. That’s a tall order.

    • Marc

      I could see it working well on limited access freeways pretty quickly. City streets are tougher and will take longer, but never say never.

      • Rob in CT

        I could see that too. But then you have the problem of the human driver having to take back over after spending however much time reading their book or what have you. Time to drive again! Get in the game!

        A lot of the really utopian predictions require the human never be needed (e.g., solving drunk driving).

    • In aviation the philosophy is “use the correct amount of automation for the situation”.

      There are times I want everything fully coupled up and there I times I need to turn all that stuff off and do it the old-fashioned way.

      • tsam

        And they’ve left you the choice. There is speculation that some of these self-driving cars wouldn’t have a pedals and a wheel. I’d be having exactly none of that mess.

        • jim, some guy in iowa

          what good would that do- in the ideal world no one would know what to do with the pedals anyway- how many people can drive stick these days?

          this whole discussion is more about age and income -and I think to a degree location- than it is technology

          • tsam

            I don’t know. I’m just a codger about stuff like that.

            • jim, some guy in iowa

              the thing that should disqualify me from these discussions is that aside from running off at the keyboard when I have time my whole gig is doing things that to a lot of people just look like grunt work- pounding nails, wrenching on machinery, working with livestock. I *like* being hands-on- but I think a lot of people don’t- or, at least, they see “hands-on” as about working more with their minds

              • tsam

                Oh so do I. And the idea that hands-on work (especially building things) isn’t mind work is pretty silly. A lot of people have that problem, and it’s a bad one. I think it feeds the anti-union stuff–where they think hotel maids deserve to be paid shit wages because they’re just changing sheets and stuff.

        • N__B

          There is speculation that some of these self-driving cars wouldn’t have a pedals

          I, too, long for a tricycle.

        • urd

          More than likely you would have to. Most legal experts think there will need to be a way for the driver to take over in case of a system problem or total failure. No insurance or car company on the planet is going to put something on the road that would expose them to the lawsuits that would come from such accidents.

          If a person screws up, the amount of money that can be gained in a lawsuit is limited (usually), but if it is now a major corporation on the line for the accident, look out.

        • so-in-so

          I suspect they will – it will always be hard to ‘tell’ a Self Driving Car “Park over there, under the tree”.

      • You and your peers, however, are highly trained to know (pretty well) what the correct amount of automation is in many (including, many extreme) situations.

        J. Random Car Driver, not so much.

      • keta

        A someone who has spent a lot of time at sea, it’s hard for me to even imagine the days when someone had to actually steer a vessel at all times. But even with all the remarkable technology, maritime law stipulates a wheel watch be maintained at all times, and only a fool would plug in some way-points and head for the bunk.

        I applaud all the advances technology has provided for increased safety, but without human vigilance ensuring these tools are functioning properly, at all times, we’re steering a course to madness.

      • Pseudonym

        Cars can pull off to the side of the road if they have an issue though.

  • trollhattan

    Automated driver assistance is already on the road and becoming more widespread and sophisticated, much of it helpful (collision avoidance ). But full autonomy seems unobtainable, given that 99.9% success leaves 0.1% of unanticipated conditions and events that require driver action. It’s the same reason passenger planes still need a flight crew.

    I recently encountered a large, busy, complex intersection where the signals were out, blinking red. Seven lanes in one direction, eight in the other including double left -turn lanes. As I sifted through figuring how to get safely through along with dozens of other drivers I pondered what the Google car would do?

    Old folks and convicted DUI drivers seem like the most likely early adapters but I suspect they’re going to be restricted to a limited set of roads.

  • pzerzan

    I keep getting tired of techies going on about the future being self driving cars and trucks. If your goal is to move tons of freight around in a fast, automated way, the technology already exists. It’s called trains. The same goes for moving people. Improving public transit is far more efficient than putting more cars on the road. That tech types would rather chase some pie in the sky dreams rather than ride a train or bus is beyond me…

    • JL

      You think tech types don’t ride trains and buses? The execs might not, but I assure you that plenty of 20-something and 30-something rank-and-file tech workers do (particularly trains).

      Sometimes public transit doesn’t go where you need to go, or doesn’t do so within a practical time frame.

      I would certainly prioritize improving public transit, but that problem will still exist no matter how much you improve public transit. I’m not the only one out there who thinks driving is about as pleasant as getting a root canal, so it’s not surprising that people are looking comprehensive alternatives. Or that people who were trained in creating computer-related products and make their living by creating computer-related products, would look to something like self-driving cars. Unfortunately, a lot of those people aren’t thinking about or accounting for the employment and other public policy considerations.

      • Philip

        You think tech types don’t ride trains and buses? The execs might not, but I assure you that plenty of 20-something and 30-something rank-and-file tech workers do (particularly trains).

        Sometimes public transit doesn’t go where you need to go, or doesn’t do so within a practical time frame.

        *waves*

        Caltrain 2x/week down to Mountain View (was 4 days a week for a while before I switched to primarily working in the city), with some biking on either end because it’s faster than Muni.

      • janitor_of_lunacy

        Where I live (rural Maryland), there are all these cute little building which used to be commuter train stations serving rail service to Baltimore. Almost no where in my county (one of the three least populated in the state) was more than five miles from a station. But the rail lines, and the stations, all closed decades ago.

    • Murc

      If your goal is to move tons of freight around in a fast, automated way, the technology already exists. It’s called trains.

      Trains are amazing for moving a kiloton of cargo.

      They’re total shit if you just need to move a ton. Or if you need to move that kiloton to a hundred diverse location.

      • urd

        Says someone with no apparent knowledge of the US rail freight system…

        • so-in-so

          Containerized or trailer on flatcar freight is basically more efficient than long haul trucking since the container or trailer takes the train to the nearest depot to the destination, then a shorter-haul truck takes it the rest of the way.

          Truckers generally hate it because long haul driving pays better. Railroads prefer (I think) unit train bulk service, but do make money on the container/trailer freight.

    • Ahenobarbus

      I keep getting tired of techies going on about the future being self driving cars and trucks. If your goal is to move tons of freight around in a fast, automated way, the technology already exists. It’s called trains.

      Wouldn’t that cost truckers their jobs as well?

  • Rob in CT

    Also,

    Where this gets “interesting” is the point at which computer-controlled cars become better then humans, but are still far from perfect. I figure people are more willing to accept human error than computer error. The computer cars will have to be many times as good as humans before they’re accepted.

    • Philip

      Where this gets “interesting” is the point at which computer-controlled cars become better then humans, but are still far from perfect. I figure people are more willing to accept human error than computer error. The computer cars will have to be many times as good as humans before they’re accepted.

      At least in good weather, that’s where we are right now. The cars are extremely cautious for exactly this reason, because the assumption is that the first time one is involved in a fatal accident, it will probably set back development decades. Because one death caused by a computer is vastly worse than 30,000 a year caused by bad human drivers.

      • Rob in CT

        No, they really aren’t that good now. They are that good in certain locations (where the maps are accurate enough) at low speeds only. So no. Maybe in 10-20 years.

        The rest, well yeah obviously.

        Insurance might be a (heh) driving force here eventually. If the numbers show the computer drivers to be superior, insurance rates will follow (there’s also other insurance question – who is the insured? but leave that for now).

        • Philip

          Highway speed driving is already being tested, just not by the ones pictured in the post. And as for locations…sure, maybe not in the nightmare of DC or Boston roads, but off the top of my head, you have Texas, Washington (state), California, and Virginia that have approved tests there. I think Michigan as well.

    • witlesschum

      Yeah. People are terrible at risk assessment, so it seems likely they’ll be way more freaked out by 1 self-driving car-caused death than 10,000 human driving car-caused deaths.

      • so-in-so

        As they are by a handful of mass transit (especially airline) casualties, but road fatalities are rarely reported outside of local news.

  • Whidby

    Self driving cars are already here.

    Most of my commute is on freeways and my car drives and stays perfectly centered in the lane and its adaptive cruise control maintains a constant distance behind the car ahead. I usually don’t have my hands on the wheel because the lane centering causes the steering wheel to twitch.

    And these roads aren’t perfect for 680, 580, and 880 between Alamo and Menlo Park in SF Bay Area.

    I do have to make steering inputs to change freeways and on surface streets, but I umderstand that Tesla has already proven concept by driving autonomously from SF to Seattle.

    People wh imagine safety issues are ignorant of how cars operate. Most modern cars are already drove by wire so there is no physical connection between steering wheel and tires and brake pedal and brakes. You are already at the mercy of electronic controls, whether you know it or not..

    I suppose one could gin up an economic argument against this fantastic technology, but that akin to fretting about buggy whip manufacturers.

    • Matt McIrvin

      Most roads in Massachusetts have barely-visible to nonexistent lane markings for about sixty percent of the year, because they get ground off by plows and road salt. The West Coast is kind of anomalous.

  • Jordan

    Self-driving cars have the potential to be *much* safer than our current system. This could save tens of thousands of lives a year, and so should be something any normal person would support.

    Unfortunately, we aren’t anywhere close to there. Even when we get closer, the likely scenario is that self-driving cars get to something like a demonstrated “parity” with human-driven ones (likely, they have fewer accidents in some areas and more in others). Once that happens, we get a big push for them (JUST AS SAFE!) as a labor-eliminating device for trucking and cabs, and people go along with it because it prevents teenage driving deaths and their commutes become nicer.

    It doesn’t mean self-driving cars as currently envisioned wouldn’t be a good thing. They’d be a great thing. But it never seems to work out that way.

    Or maybe I’m just being grumpy today.

  • witlesschum

    Seems to me a downside of self-driving cars is that now people will be expected to work during their commutes, destroying the only good thing about anyone’s commute.

    • Yep. This is exactly what will happen. All that supposed freed up time for the middle class just means more demands from employers.

      • brugroffil

        How well do employers currently monopolize public transit-using employees? How would the result be any different in your preferred model of more density and better mass transit?

      • Matt McIrvin

        We can test this by looking at hours worked vs. distance from the office, can’t we? Do people who live 10 minutes from work as opposed to two hours just work more hours to make up the gained time, or do they have any more free time?

      • UserGoogol

        This doesn’t really make any sense. It’s not like most people are going straight from bed to work to bed, there’s plenty of hours available for employers to take from workers that they have not yet done. The amount of hours workers work is based on negotiations beween workers and employers (be it individually, collectively, market-based, government-based, etc). Employers have a certain amount of upper hand in the situation, but they don’t completely control the situation. Presumably, if more free time is made available that changes the dynamic and some of that time is taken by employers, but it’s absurd to think all of it would be. When people are working 112 hour weeks you can say that, but not now.

        • urd

          You must not live in the Bay Area or LA. While it isn’t literally as bad as people getting home from work and then going to bed, I do know of enough people with horrid commutes that they come real close to it, especially once they take care of evening tasks that have to be done each night.

    • Jordan

      ahh, that is another downside :(.

    • Whidby

      How is that a bad thing? I get caught up on my emails during the drive in so I am inbox zero when I step into my office.

      I prefer it to working at home where I have to spend the first hour wading through the the overnights.

      • tsam

        Some of us like working a 40 hour work week because families and responsibilities at home. You know, silly shit like that. Some of us are sick and tired of the expectations of incompetent managers who force their employees to work while driving to work, or at home, or on weekends, etc…

        • Whidby

          I guess I don’t understand this antagonistic stance that som workers have with their employers.

          It seems unhealthy

          Why not just quit?

          • Quit and do what?

            • N__B

              “I don’t know, man. Sit around and groove on the rubble.”

              I owe a beer to whomever can identify the quote…

              • The Dark God of Time

                Gabe Kaplan as Abbie Hoffman from “Richard Nixon, Superstar”

                “we’d all be a lot happier if he did to her what he was doing to the country”.

                • N__B

                  I owe you a beer.

                  “Late at night…Whittier, California…a lone star shines brightly overhead.”

          • Jordan

            classic whidby

            • He’s off to Florida for a wild spring break!

              • Jordan

                dude is almost certainly an MRA frat boy, right?

          • tsam

            Why not just treat your employees with respect and dignity, and run and staff your office correctly so that all the extra work isn’t needed?

            Or a person could quit and do the same thing somewhere else. I guess that’s a viable solution.

          • witlesschum

            Sometimes, it’s kinder to assume a person is trolling than that they’re being serious…

          • Because jobs are so easy to get these days?

            Good luck if you’re over 50 40 because you’ll probably only be able to get a low-wage service sector gig.

          • DrS

            Are you sure your name isn’t actually Willard?

          • tsam

            And again–if you have to work at home at night, on the weekends, while you’re fucking DRIVING, then your manager is incompetent. Sometimes things happen and people have to put a little extra time in, but that should be the exception, not the rule.

            I own my business. My employees are expected to go home after work and do something besides work for me. I don’t ever want them to feel obligated to work 60 hours a week just to meet my expectations. That would make me a terrible manager.

            • N__B

              Seconded.

        • Sebastian_h

          Does your employer regularly force people who commute on trains or subways to work for them while on the train or subway every day? They might for work travel, but I’ve never heard of it for the commute.

      • Philip

        Because companies shouldn’t be able to take over any more of their employees’ lives than they’ve already managed, and the chances of regulatory intervention in most of the US are just about 0.

      • witlesschum

        Because employers getting control of more and more of our time is a bad thing, especially when it’s not accompanied by proportional increases in compensation.

        Maybe it’ll work out fine for you, but it seems more likely you’ll be required to still do the emailing at home because you’ll be forced to do something else on the way to work in the self-driving car.

        • Jordan

          Right about lots of this.

          I would point out that shift workers who have to drive to work, however, probably would benefit quite a bit from this type of ai-commute. They could catch up on sleep, or whatever else they need too.

          white collar workers and whatnot would surely suffer.

        • Matt McIrvin

          See, to me, the time I spend driving in rush-hour traffic is already time that my employer is forcing me to do something incredibly unpleasant. I realize that this won’t apply to everybody, but I find it far more unpleasant than actually working.

          I guess some people see that activity as their own private time, and I guess listening to the radio is all right, but otherwise it’s just incredibly horrible, like being forced into a continuous fight-or-flight response with honking screaming morons for an hour and a half.

          • witlesschum

            I listen to podcasts or audiobooks on my commute, so it’s pretty much recreation time for me. Not as relaxing as taking a bus or walking to work, but I’m comparatively strange in that I live in the city and work in BFE.

            • Matt McIrvin

              Ah, the reverse commute! I used to have that. It’s nice if you can get it.

    • xq

      I hardly ever see people working on the bus/train. And many people can’t work during their commutes because their jobs require being at a particular location.

      • brugroffil

        Yeah, it clearly isn’t a universal thing. Anyone with a service sector job won’t be working during their commute.

  • mds

    I guess I’m forced to agree that the infrastructure problems aren’t insurmountable. As long as all the people who are currently so enthusiastic about this have access to well-maintained private toll roads with adequately-funded tech support, there won’t be any problem.

    • tsam

      The infrastructure problems aren’t insurmountable, no, but the ecological problems of focusing development on individual cars, rather than mass transit seems like it’s insurmountable. I get the car thing, I love mine, but we’re reaching a point where that model of transportation is endangering the planet.

      Also hamburgers. But mostly cars.

      • so-in-so

        Doesn’t this depend on how they are used? I see comments about increased utilization – if your commuter ride is going to earn it’s keep during the time you are at work, you’d better be in a pretty dense area. It doesn’t make sense to have the vehicle drive another half hour for it’s next ride.

        • tsam

          I’m clumsily advocating for mass transit in that comment, rather than focusing on individual cars–electric or otherwise.

  • Patrick

    One things that struck me is how the article called out American roads as uniquely poorly maintained and differing by jurisdiction vs. the European car makers expectations. But is that really true vs more than just in Germany?

    Are there EU wide standards that are more adhered to than US MUTCD standards? Straying from those will get your dept of transportation sued pretty quick by the first guy who gets T-Boned at a non-uniform intersection.

    And are roads/markings really worse maintained than those in middle of nowhere Bulgaria?

    • Warren Terra

      I don’t know what the roads are like in most of the US, let alone Bulgaria. And I don’t know what these robots actually require. But I can tell you that driving on I5 out of Los Angeles towards the Burbank Airport in a foggy twilit morning can be extremely nerve-wracking because the worn, faded lane markers just don’t reflect your headlights (heaven knows when they were last repainted). And that’s one of the busiest, best known roads in one of the biggest cities in the world.

      • Has the USA heard of cats’ eyes?

        • Rob in CT

          Yes, but you don’t see them up here in NE much. I think because they get killed by snowplows pretty quickly. I’ve seen ’em down South.

          • Percy Shaw’s original design of 1936 included a snowplough-resistant heavy steel shell for the rubber insert carrying the reflector. It snows in Yorkshire, thar knows. Where the US went wrong was in going for the rubbishy Bott’s dot.

            • Rob in CT

              Well, assuming the $$ works out I don’t know why we don’t use them then.

        • It took a surprisingly long time for them to be developed (as they are now) to the point where they could withstand winters in the Northeast (snow plows…).

          Why even the much older version shouldn’t work on I-5, I have no idea.

          • Pseudonym

            I think I remember seeing recessed retroreflectors on I-5 up around Mt. Shasta. That gets slightly less traffic than in LA though.

        • tsam

          Has the USA heard of cats’ eyes?

          Yes. When we were blackout driving at night in the Army, we had to follow those little tiny cat’s eyes on the tail lights.

  • DrS

    Wouldn’t it be simpler just to lay down tracks? Seems like it solves many of the big technological issues.

    • Warren Terra

      You are joking, right?

      • DrS

        Mostly.

        I’ll note, however, that the cars on Disneyland’s autopia manage to stay on the track.

  • Grumpy

    Erik, you keep saying that the proponents of self-driving cars have to answer for all the jobs lost as a result (truckers, etc.) but wouldn’t mass transit lead to loss of jobs? Mass transit is more efficient and even easier to automate than cars, right? So why is mass transit defensible from the blue-collar perspective?

    • Rob in CT

      I’m sure if Erik was around when the railroads ditched steam engines in favor of diesels he’d have been pissed.

      • sparks

        Why? It didn’t change anything as far as who worked on the trains. There were still engineers, firemen, brakemen, and conductors. Ditching the caboose took away more jobs than diesels.

        • Rob in CT

          Diesels require a lot less maintenance.

          • so-in-so

            They require fewer crews if more than one locomotive is needed (steam engines required a crew for each, form the 1940’s on diesels would be controlled electronically from the lead unit).

            Diesels were arguably better for the environment. A diesel electric uses 30% of the BTUs in it’s fuel for motion. The Steam engines used about 7% or the BTUs with the rest as waste.

            • Jackdaw

              There was also no more need for a fireman (whose job was literally to keep the fire of the steam engine stoked).

              • so-in-so

                Yet the unions managed to keep a fireman (as an assistant engineer) on trains until the 1980’s in most cases. Then they removed the caboose from freight trains and the conductor became the second person in the locomotive cab.

        • Pseudonym

          Diesels don’t require the same kind of firemen that coal-fired steam engines do.

    • Wapiti

      Doesn’t mass transit lead to more jobs? If 1000 commuters take the train instead of 900 cars into the city, that means we still have 1000 employed commuters + some number of transit workers.

      But I guess we have fewer people working the gas pumps, fewer car mechanics for routine maintenance.

      • Grumpy

        I misunderstood you for a second. Yes, mass transit would employ workers initially, but I (being totally ignorant, admittedly) assumed that work in mass transit is going to be quickly automated. I don’t see it being better than cars for workers, that’s for sure, and that’s Erik’s whole case in favor of cars.

  • I view automated cars as more or less inevitable in our current legal climate. What will drive the push for automation is the fascination of the American public with technological innovation shiny gadgets, insurance companies and the NHTSA pushing more safety gear. Right now, cars can use a GPS map to navigate roads, but when the information loop is closed and a few million cars report their GPS positions back to a central navigation database, so that they can all work from an up to the minute (or up-to-the-last-time-a-car-drove-that-route) map which will even include the locations of potholes, roadkill and highway construction, the driving experience will be much better.
    If Tesla motors isn’t already building such a map, I predict they will start within the next year or two, and when their new model C cars hit the road, there will be an order of magnitude more electric cars on the street.

    To be clear, automated cars have all the drawbacks that Erik and others have already mentioned, but even without better infrastructure, automated cars (but maybe not autonomous ones) are on their way.

    • Philip

      That map, more or less, already exists, although it relies on human reporting right now. That’s the selling point for Waze over other smartphone GPS systems.

      • But requiring human reporting hamstrings it. When data from the cars radar collision avoidance, temperature sensors, windshield wipers, GPS and brake pedals is automatically uploaded, the car will know, to “watch for deer at dusk on this stretch of highway 51” or that “this section of main street gets slippery when it snow, better slow down”. That is what will take automated driving to the next level.

        • Philip

          Sorry, yeah, what I meant was that nothing new has to be created. The map and reporting system exists. It just needs to be extended and have cars’ sensors plugged into it.

  • Wapiti

    A lot of commenters speak of the tens of thousands of lives that might be saved… My cynical view is that, like gun deaths, car deaths are just part of what some subset of Americans are willing to spend for their lifestyle.

    If car deaths need to be curbed, take the worst 10% of the drivers off the road. DUI? Get your licence back in ten years. Reckless driving? Get your licence back in five years. Past 72 years old? Eye and reaction exam every three years.

    Unfortunately, our cities/suburbs are so twisted by the auto that many people can’t handle getting to work without one.

    • Add the cell phone yappers and texters to your list.

    • NonyNony

      If car deaths need to be curbed, take the worst 10% of the drivers off the road.

      I suspect if we spent more money on our infrastructure we could curb car deaths without taking anyone off the road. A lot of accidents happen because of poorly designed roadways that are still in use despite engineers knowing that the designs from the 1950s are deathtraps. A lot of other accidents occur because of poor maintenance on the roadways. And still others because we like to do things on the cheap.

      • Wapiti

        Just a gut feel of mine – when I lived in the SF Bay area I wondered if some of the highways and merge lanes were designed for 55 mph and were later re-signed for 65 without any change to the merge distances and sight distances.

  • The whole problem I have with robot chariots is that it assumes a future where we’re not going to be primarily preoccupied with survival. Maybe robot chariots will have a place in the Elysium communities that the uber rich create for themselves, but the rest of the world will be too preoccupied with trying to find food, water, and a cool place out of the heat to be driven around by robots.

    • Wapiti

      I’d agree. The robot-driven car taking a passenger 20-30 miles from home to the place of work is frankly not advancing the bigger problem of reducing carbon emissions.

  • Gregor Sansa

    The question of whether it will happen is independent of whether it’s a good thing.

    The question of whether it’s a good thing in terms of traffic fatalities is independent of whether it’s a good thing in terms of jobs.

    The question of whether it’s a good thing in terms of jobs is NOT independent of many other questions of economic organization that will be being addressed with increasing urgency on all sides whether or not this particular technological dream becomes a reality.

    Erik is justifiably angry at the capitalist class, but mixing up these questions goes past being good trolling, and becomes just muddy-headed thinking.

    • Ultimately it comes down to this: give me three scenarios where self-driving technology can be purposely misused, and I will consider the problem at least somewhat thought out.

      Anything that can be misused, will be misused. The techno-utopians are all about the good, but one thing I never hear about is serious discussion about how a bad actor can misuse the technology.

      • Philip

        Then you’re not paying attention. There’s been a lot of talk about the potential security problems in these cars, for example. The answers so far, based on the total lack of security in non-self-driving-cars, are not good. But people have been talking about it.

        • Oh, I’ve been paying attention. I meant the comment in general terms. I think there a many ways in which this technology can be misused, and not the least the affect it will have on privacy and the freedom to travel and congregate. Just wait until your robot chariot refuses to take you to that Occupy rally and logs that you asked it to take you there.

          I’m just saying that you almost never see it discussed in comment threads like these. You tell me the technology is awesome? Great! Now tell me some of the downsides and I’ll consider that you actually thought about it :-)

          • so-in-so

            Your car already has a computer and probably a GPS. If it has internet access and/or one of the recent theft prevention packages your scenario is theoretically possible NOW.
            Do you carry a cell phone with GPS?

            • Philip

              In (most) cars, the core functions aren’t currently accessible from the networked bits, though. The head unit has no control over brakes, steering, and what have you. That’s inherently different in a self-driving car.

              • so-in-so

                Actually there were news reports about hacking current vehicles that indicated the core functions WERE accessible. They certainly are on the theft prevention systems. Certainly they could (and should) be separate. For that matter, the self-driving features only need access to the GPS unless some other networked function is used, so the issue is the same.

                • Philip

                  I said most :) The Jeep problems were much worse than, e.g., the Nissan problems, and most cars’ architecture is closer to the Nissans’ than the Jeeps’.

                  I suspect even just being able to hit the GPS and real-time mapping systems will be enough to cause problems. It’s a complicated system, and there are always holes in attack surfaces like that.

            • Yes. But I also have the option to leave my cell phone home. You could also say that I will always have the option to walk. But how long until walking is considered suspicous in a world where you could just as easily take a robot car door to door?

              • so-in-so

                Yes, and not having your cell phone with the GPS enabled could be considered suspicious as well. There are certainly negatives to self driving cars. This one seems a bit like the Y2K worries about “what if firetrucks have some system that would let the run when it thinks they have missed too many maintenance cycles?” That one is a little worse, since nobody had any evidence that such a system existed at all, but this case assumes dystopian politics (which will have much more serious effects regardless of self driving technology).

            • Ha! Let’s see them hack my 1957 DeSoto!

              Jurassic technology for the win!

  • Everybody is missing one huge prospective benefit: the enormous reduction in the number of vehicles, and in the need to park them.

    The argument goes that a self-driving car breaks the affective link between the driver and his (let’s face it, this is mainly a boy thing) motorized steed. One self-driving car is as good as another. So self-driving cars will be shared through Uber-style apps. Guessing, advocates put the reduction in the number of vehicles at three-quarters. That will liberate huge areas of parking space, especially in city centres: this can be cashed as a better environment or for more development. Garages will become dens.

    What self-driving won’t help at all with is congestion. The argument for congestion pricing on low-occupancy vehicles and better public transport stays intact.

    • And when this need for speed is no longer fulfilled, what will the boys do then?

      • I thought they were already sending each other erotic selfies.

    • Warren Terra

      the enormous reduction in the number of vehicles

      I’m not entirely clear on this argument – what proportion of all cars are in use during the daily commute, at the same time? More efficient use of existing cars by contracting rides (either with a robot or a human driver) instead of owning your own car will, on its own, only reduce the number of cars on the road by some proportion of those cars not in use at the same time every day. I’d guess that number is relatively limited.

      On the other hand: if people really get out of the habit of driving their own car there is a lot of potential not in car shares but in ride shares. Uber is already introducing this (though, I understand, not offering the user huge discounts) when it has two different users wanting to take basically the same trip at basically the same time.

      and in the need to park them.

      This I agree with, especially because if people use hired cars when shopping it could greatly reduce the vast parking areas of retailers. But, for the commuters’ cars, is this true? Unless people are sharing them and so greatly reducing their numbers (which I’m skeptical above), they’re still going to have to mostly wait until the end of the workday – unless they drive all the way home and later back to work again!

      • See my comment above. Give me three ways in which this technology can be purposely misused.

        • Rob in CT

          Hack to cause accidents (anything from hacker shits & giggles to terrorism).
          Hack to steal stuff (re-direct delivery van, order car to unlock itself so you can take it, etc).
          Like Rick Perry, I was sure I had a #3 but can’t seem to find it in my brain.

          • so-in-so

            Both apply to the current generation of driver cars with internet access (yes, that’s not all cars).

          • And that’s just the stuff you came up with off the top of your head. Now consider what a team of highly motivated and/or financed hackers could come up with. What could the NSA come up with? How much money has Apple spent on securing their cell phones? How did long did it take a dedicated group to help the feds to crack it?

            • DrS

              Although with that last, my understanding is that the phone was on an older version of the lock than is currently available and the murderer didn’t take any of the extra available security precautions that were included.

  • So we won’t even be able to get shitty jobs driving for Uber any more?

    Gosh I can hardly wait.

    I suppose there’s still Starbucks, but Silicon Valley is probably already working on the Baristamatic 5000 cappuccino-bot.

    • Philip

      New York will have the Baristo-5K long before the West Coast, where people expect and demand the opportunity to have an awkward 30 second conversation with the barista every morning for some reason.

    • Pseudonym

      We already had those in the kitchenettes back when I worked at [Silicon Valley company now working on driverless cars].

      • Philip

        They make pretty iffy coffee though

        • Pseudonym

          A lot fewer milligrams of pretentiousness than Philz however, for those of us who have to watch our daily intake.

          Our first purchase as a startup was a (manual) espresso machine, for what that’s worth.

  • leftwingfox

    Reposting CGP Grey’s “Humans Need Not Apply” video, which I linked August of 2014:

    https://www.youtube.com/watch?v=7Pq-S557XQU

  • N__B

    The technology described is long since obsolete, but the human issues are not: Aramis.

  • Pseudonym

    I’m sure when the silent Luddite anti-technology majority of America finally puts its foot down and stops compromising with major political parties there will finally be plenty of well-paying jobs that don’t require a graduate school education.

  • wengler

    So many people think there are a large amount of technological leaps to be made. There are not. It is all about reliability or affordability.

    The problem with autonomous vehicles right now and into the future is mostly legal. Who is responsible when autonomous vehicles are the most convenient way to blow up town square? Imagine a fleet of autonomous trucks all with OKC-sized ammonium nitrate bombs in them all timed to go off at the same time.

    • so-in-so

      If the absence of self-driving vehicles is the only thing stopping this scenario, we are truly screwed. The last time anything close happened it was a rental truck. If stopping google is the only thing between us and this point, then I’m sure somebody will be jacking some UPS trucks real soon…

  • brugroffil

    What are the legal/insurance ramifications when two self-driving vehicles hit each other? How do you “prove” who is at fault? Does the human “driver” get a citation?

    • Rob in CT

      Obviously you don’t give the human non-driver a citation. Proving who is at fault is probably not a lot different from today, or maybe easier (review data from both cars). Today how is it done? The cops talk to the drivers of the cars and maybe some people who saw it. Also: look at the cars. Sometimes the damage largely tells the tale.

      Insurance would probably be to the car company not the driver.

      This is for fully-automated cars. In my lifetime, we’re likely in the human-augmented zone and insurance is pretty much the same as today.

  • MDrew

    So if it didn’t maybe threaten trucking jobs, would you be okay with pursuing the technology because of the potential to save thousands of lives?

    The reality is that both are speculative. It’s entirely possible there will be reasons for trucking companies not to adopt the technology – including possibly a new law passed at the urging of the Teamsters saying commercial applications are not allowed, or other feasibility issues. One simply never knows.

    One also never knows how likely it is that many lives are eventually saved by the technology. So why not move forward and find out? A danger of not pursuing the consumer transit side of the technology is that we could give up whatever upside there is there while failing to prevent adoption of the technology by shipping companies.

    From a labor perspective, or for that matter a societal perspective, how do we ex-ante distinguish technology to pursue from technology to shut down? Just as Erik? Or just shut all new avenues down – we’re where we need to be (or a bit past it); we need to stop in our tracks?

  • Barry Freed

    Self driving cars will not happen for at least 30 or more years. The problems are pretty severe and are not getting solved. They have a problem with bad weather and mixed non-standard signage and they’re easily capable of being pranked. Like if someone steps out into traffic as if they are going to cross the road but don’t actually cross just to make the car stop. People are going to fuck with these things because that’s what people do. They’re really bad at predicting the behavior of people, something people do easily without even thinking about it. I don’t think any of these problems are anywhere near to being solved.

  • Even if they start selling these tomorrow, it would take quite a few years before there was a large percentage of them in the fleet.

    The average age of the US car fleet is 11.5 years (I looked it up).

    My car is 12 years old. My wife’s car is 7 years old. We have no planes to replace either one any time soon.

    • Denverite

      We have no planes to replace either one any time soon.

      Teeheehee.

    • Hogan

      My car is 12 years old. My wife’s car is 7 years old. We have no planes to replace either one any time soon.

      And the planes probably wouldn’t fit in your garage anyway.

    • Warren Terra

      My car is 12 years old. My wife’s car is 7 years old. We have no planes to replace either one any time soon.

      I think it’s a shame people are harping on the “planes” typo when the obvious reaction is profound relief that Major Kong and his 12-year-old car have no immediate plans to replace either the Major’s wife or her 7-year-old car.

  • Pseudonym

    It’s reasonable to want truckers to be able to keep their jobs and to find that more important than autonomous driving (there’s no reason that those technologies couldn’t be used to augment manual driving for additional safety, as indeed they are being deployed in vehicles today). What’s the limiting principle on opposing disruptive automation though? Furthermore, how is the current model of long-haul trucking environmentally sustainable in the first place? Maybe the jobs will be in constructing and servicing wind and solar farms to power the autonomous electric trucks. But the kind of massive infrastructure investment and technological change that is going to be necessary to, say, keep Florida and Bangladesh above sea level is incompatible with Erik’s approach to the future if it’s intended as an actual plan rather than an emotional reaction.

  • Pseudonym

    The other big question I see is how autonomous cars will change living patterns. Building better road infrastructure in the US seems to lead to people living further out in the suburbs away from work. Is there some constant amount of commute pain that people will put up with, so that if self-driving cars ease the commute, people will spend four hours a day riding in their cars instead of two hours driving? I think autonomous vehicles are an incredibly interesting engineering problem but am skeptical that they are going to usher in any significant improvements in quality of life on their own, absent some more enlightened housing, urban planning policy, and mass transit policies that we should be seeking regardless.

    • Warren Terra

      For a class of worker who can get their work done in the back of a car, given decent noise cancellation and internet, this could easily increase the time and distance they’re willing to spend in a commute. In that sense self-driving cars could easily increase the stratification of our society, as the boss or even the midlevel executive spends two or three hours each day teleconferencing in their luxury robot-chauffeured rolling office to and from their rural estate.

  • Pingback: Non-Sequiturs: 04.04.16 | Law Deluxe()

  • Pingback: Osibogun and Partners - Law Firm()

It is main inner container footer text