Google and the self-driving delusion

“This year Audi announced its presence [in self-driving vehicles] by sending journalists on Monday to drive the 550 miles from Silicon Valley to Las Vegas in an experimental A7 sedan equipped with what the German car maker calls its ‘piloted driving’ technology,” Holman W. Jenkins, Jr. writes for The Wall Street Journal. “Note that Audi is not one of the nutters. Its engineers go out of their way to distinguish themselves from Google ’s engineers, whose latest driverless cars come without steering wheels. Audi says definitively: ‘Audi will never build robot cars.'”

“Its journalist test drivers, who underwent special certification, were supposed to be able to hand control over to a computer when creeping along in stop-and-go traffic or when droning down the superslab. To maintain a fixed freeway speed, the computer was even supposed to be able to cope with slow traffic by changing lanes and passing without assistance from the driver,” Jenkins, Jr. writes. “But here’s a question: Why would a driver activate such a system except to turn his attention elsewhere? Yet Audi’s piloted-driving system expects the driver to remain alert enough to respond within a few seconds if the computer decides to hand back control to him. If the driver isn’t ready, reportedly the car will turn on its flashers and find a way quickly to bring itself to a stop.”

“Is this not crazy, the equivalent of an airplane computer dumping the job of flying back in the pilot’s lap just at the moment when 10 things are going wrong at once?” Jenkins, Jr. asks. “Which raises a question: What is Google up to with its driverless car, which it diligently subjects to fake real-world tests, though it will never be a real-world product?”

“Google has lately accrued some pessimists for 2015, analysts who believe the company is approaching a growth impasse, and will have to start becoming more shareholder friendly,” Jenkins, Jr. writes. “An obvious place to start cutting might be Google’s driverless-car efforts. But this would be to misunderstand the point of an elaborate charade. Google’s driverless car has become a branding exercise, a ticket to free media, a way to market Google software to auto consumers for onboard infotainment systems. Never mind the false expectations it raises in the driving public. Never mind the pressure it puts on real auto makers to pay lip service to autonomous driving they won’t be able to deliver.”

Read more in the full article here.

MacDailyNews Take: It’s nice to see someone else clearly understanding that Our Lady of Perpetual Beta’s pie-in-the-sky products aren’t innovation, they’re just mental-masturbatory marketing ploys.


  1. I just can’t see completely driverless cars being something that the public at large will want to just transition to in one go. It seems massively more sensible to gradual improve the assistive devices available to vehicles. As the technology improves and people get used to it we may get to a point where cars are essentially completely automated but Google’s method seems very definitive.
    Also, we recently jiggled cars around in the driveway to get them all in over christmas, I can’t see an automated car understanding the need to move a car over a few inches or overhand an area of grass etc.

    1. If you think about it a little more, a self-driving car needn’t only be driven while you’re inside it. You could send it off to park round the corner rather than not block your other car in the drive. Then when you want to drive it, you can call it up. Similarly when you get to your destination, you could also elect to make it park in a cheaper car park a little further out of town and then collect you again once you’re done.

      Alternatively, a really smart car would be able to access your diary and those of the others in your house, so that if the last car to get home is blocking others in the drive, they could decide to automatically rearrange themselves so that the car which will need to be used first is unobstructed in the morning.

      1. The point is, going from what we have now to Google’s entirely automated self driving only option is a huge leap. I just don’t seeing people being confident enough with it for it to take off. Self parking, lane assist, etc are all things that can be added gradually and get better and better without forcing people to make a huge leap of faith. In time we may end up with having cars that will drive themselves, but Google’s way doesn’t seem like one that will work to me.

      2. Jesus it sounds like you actually believe this will happen. On a serious note as wonderful as this fantasy sounds I’m sure the last thing we need is thousands of driverless cars driving around driverless and passenger less simply to find cheap parking places or for some other convenience. Perhaps we should do away with the passenger cabin altogether and all head for the driverless public transport.

    2. This is a very sensible approach but of course far too below the marketing radar for the likes of Google to be interested in. I have seen the Audi solution which appears far more logical and advanced than Googles yellow submarine approach. However this writer is spot on in relating the total contradictions even in Audis approach. We are presently in a World where state of the art trains take known corners at excessive speeds and derail despite all their safety devices so how the hell could anyone in the foreseeable future believe we can rely on autonomous cars even with human presense just like those crashing trains attempting immeasurably more complex manoeuvres.
      Worse still Mercedes and others are contemplating cars with inward facing seats based no doubt on the same logic that early tube/subway trains did not need windows. That was soon in practice shown to be totally incorect as will the concept that passengers could ever feel comfortable with their backs to the direction of travel as their autonomous car speeds down the road while they chat without a care in the world. Hey why not revive that train logic, fill in those windows and let them watch films at 70 mph on the motorway.

  2. I would never buy a car without a steering wheel…but I can see the use of a driverless car as part of a public transportation/city transportation system. I’m no fan of Google- but I believe this technology will be used sooner than later.

    1. I agree with both points. I don’t know how it would work in the real world, but I would rather a drunk driver hand control of his car over to a computer that try to drive it himself.

  3. Google is like the guy at school who was really good at football or was the fastest runner and everyone thought he was cool but then when you leave school and grow older you work out he was a bit stupid and he still works at McDonald’s cause everything else he tries he does really badly.

    1. Exactly! Two of the star football players on our high school team, whom everybody thought were so cool ended up as a gas station attendant and a furniture mover.

      Endlessly churning out “cool” new beta products that never really accomplish anything is not leadership in innovation.

      1. Three words Zeke:

        Traumatic Brain Injury.

        Some of these star athletes may have had the brains when they entered high school, but not by the time they left high school.

        iCal this: 20 years – no more H.S. Football. 25 years – College. 30 – NFL.

        Sorry America. It’s future true.

        1. More like in 20 years we’ll see less offensive weapons in football, like less elaborate padding and helmets. These are billed as being defensive items that prevent injury. In truth, they are offensive weapons that allow vicious hits while protecting the player who delivers them. For example if players wore the simple leather helmets of 1930 there would be way fewer helmet to helmet hits, and you wouldn’t need rules against them with the penalties and fines that are attached. They don’t have anywhere near the same levels of these problems in soccer and rugby. The difference is helmets and pads.

  4. I do NOT like Google and agree that most anything they are working on today other than capturing your data through search technology is a complete waste of their time. Disclaimer: I prefer duckduckgo.

    However, I’m fairly certain Orville and Wilbur Wright had no idea a plane like the Airbus A380, the size of an office building that could fly nearly 10,000 miles would be a result of their early work.

    Point is… a driverless robotic car WILL happen someday. Not in my lifetime I’m afraid, and maybe not even in my children’s. But it WILL happen. And this early work, albeit crude and “masturbatory” will help build the foundation.

    1. Autonomous-capable cars may come sooner than you expect.

      Maximizing safety, fuel efficiency, speed, traffic, and personalized transportation (rather than public transportation) is a such a strong incentive that we’ll soon see significant federal interest in promoting such programs. Obviously, it will require that such vehicles be networked for maximum safety, speed, and efficiency, but they’re already offering cars with gps, avoidance sensors, and wifi. It’ll be a shame to lose personal driving control, and non-autonomous vehicles will have to be decentivized and live by even more strict rules until the transition is complete, but the potential is hard to disagree with, once the technology gets there. I suspect there will first be autonomous-only lanes for fast, efficient travel that will increase and gradually crowd out the restricted individual drivers.

      But imagine ultimately being able to autonomously travel in your own vehicle between destinations at speeds approaching twice the current levels, without stop-and-go, or traffic jams, or drunk, sleepy, senile or otherwise-incapable drivers, and all while saving energy at the same time.

      Once upon a time it was inconceivable (except to a few dreamers) that the automobile would ever completely supplant the horse, but look how quickly the transition occurred.

    2. No one is saying on won’t happen. Many manufactures are working on driverless cars. What they don’t agree with Google on is no manual override. Check out the Mercedes one, I’d drive in that, I won’t go near Google ones

      1. I think you’re missing the point. There WILL be driverless cars WITHOUT manual override one day. Somebody has to do early pioneer work for this to come to fruition. Of course, nobody is arguing that you should have no manual override today or even in the next decade. But eventually you will need and want it.

        1. “There WILL be driverless cars WITHOUT manual override one day.”

          They’re here now. It’s called light rail, and it was designed to be driverless. However, due to safety concerns systems like BART and MAX are required to have totally redundant human drivers. I suspect it will be the same with cars and trucks.

  5. Nothing will happen with driverless cars until this question is answered: When a driverless car runs over a three year old who pays? The owner of the car says: google promised me it would work. Google says the owner has the ultimate responsibility. The owner says to google: I didn’t run over the child you system caused the car to run over the child.

    It is equivalent to General Motors making cars with no brakes and telling customers you can now go faster.

    As far as the future goes, maybe driver assist systems could make parking, commuting and long range travel safer but driverless systems will only work in closed systems like the Atlanta airport. Someplace where EVERYTHING is controlled.

    1. A self-driving car hitting a pedestrian is no different than a manually-driven car hitting a pedestrian after running a green light. If a pedestrian runs into the path of a vehicle illegally (i.e. outside the marked street crossings, when the traffic light is red, etc), the driver of the manually-driven vehicle would fare no better than the self-driven car, and the fault would still the pedestrian’s.

      There is no doubt that self-driven cars will have far better prediction and response accuracy in traffic than the best skilled drivers of today. That only makes sense as the image processing and analysis software becomes more efficient and accurate. It is just the matter of time.

  6. Ehhhhh, idunno. I think if I’m gonna spend more than $50K on a car it’s because *I* actually want to control the thing: curvy roads, burnouts, donuts, all that fun goofy stuff. If I want to be driven I’ll take a bus, subway, cab or a ride in the cop car because I was misbehaving in my non-Google car.

    A guy from Allstate was talking yesterday about future liabilities as a result of Google cars gone wild. Not entirely beyond the realm of possibility. I’ll stick with my trusty, sporty Ford Pinto for now.

  7. And when it’s an icy day and your cameras are iced up, does the car go or force you to defrost first? When the snow keeps coming down, does the car pull over (or stop dead) when it can’t see? When you hit black ice at the corner, does the car get you out of the spin before the back end hits the group of kids waiting for the school bus?

    Driving down a nice clean dry road in a well-mapped area like San Francisco, in a car that gets serious maintenance every day, is not the real world. The real world is when your cameras have gotten covered in dust and your transponders are cranky because you haven’t been able to get to the dealer for that $500 maintenance. The real world is when there’s a puddle under that bridge that you can’t see because it’s in shade. The real world is when some clever North Korean hacker drives your own car off the bridge because he didn’t like your jokes. The real world is messy.

    We’ll all be dead and in our graves before this really works. You can fool Wall Street, but you can’t fool Mr. Murphy.

      1. >>These cars will be able to see through snow with radar, react to black ice faster than you ever could

        Yeah sure. In 50 years maybe. A lifetime of high-end software development makes me pretty confident that my guess isn’t too far off.

        1. My guess is in 10 years at the latest. Musk says 5 for Tesla which probably means 3. Mercedes is also making a big push:

          I see the traditional car becoming obsolete within 20 years. In addition, there will be less cars on the road in the future as there is today. If drones become accepted then people will just have their Taco Bell delivered via drone. When a people mover is needed they will just Uber one up.

          There still will be a market where individuals or families purchase vehicles, but the designs/function of these vehicles will be radically different from what they are today. They will become living rooms and offices. Some cars might even have beds. There will also be autonomous party buses, equipped with bars, etc which will transfer Friday nighters from the suburbs to the cities. There might even be ones equipped with pools and lounge chairs which will transfer patrons from LA to Vegas, etc.

  8. My three year old Chrysler 300 has adaptive cruise control (ACC), and will speed up/slow the car as traffic speed fluctuates. I drove from Sacramento to Seattle last year and never touched the brakes except to leave the freeway.

    New cars today will track the painted lines on the road, and steer your car to stay within the lines, while warning you that you’re wandering out of the marked lane.

    New Mazdas will actually stop the car if traffic stops.

    My GPS navigation system even knows the posted speed limit.

    So, it’s not hard to imagine the day of the rudimentary self-driving car is nearly here. Add a few motion sensors to the bumper (which my Chrysler already has), blind spot monitoring (which I also have), front/rear cameras (I have rear camera), and software tie-ins to the navigation system, and you’re there.

    I agree people will not trust this technology to start, as I didn’t trust cruise control on my ’83 Toyota Cressida, but I got used to it, and so will many as the technology becomes more mainstream.

    I think we’re closer to this technically than you imagine. The psychological barrier will take a bit longer, but ubiquitous availability in a few years will close the deal.

    1. The things you describe are absolute child’s play compared to what a real self-driving car will have to do. To wit:

      Painted lines on the road. Don’t complain about the taxes you’ll have to pay to keep ’em nice and bright. And the embedded wires to keep the ice off the several hundred miles of road around wherever you live.

      Does your GPS work well in canyons? Mine doesn’t. We’ve got an awful lot of canyons out here in the west.

      I can keep them coming, but it appears people have got serious woodies about this imaginary tech. Keep the viagra coming, cause you’re going to need to sustain those woodies for a long, long time.

      1. I watched a car auto drive at CES online today, even putting on the turn signal, stopping at a corner, and proceeding to turn and proceed to its destination. The car had front and rear cameras, which were visible in the center console display.

        This is moving quickly to becoming mainstream.

Reader Feedback

This site uses Akismet to reduce spam. Learn how your comment data is processed.