What Ford’s New Guy Said about the Future of Self-Driving Cars

You get to own and steer a car for a while longer, if you insist.

Ford Motor Co.’s new CEO, Jim Hackett, is approaching the end of his self-imposed 100-day review of the company’s operations. And self-driving cars is the big thing.

He took over after Mark Fields was sacked on around May 21 for a lack of strategy, a public tiff with the Trump administration, declining US sales and market share, production cuts, layoffs, and most unforgivably, the share price.

The stock sagged nearly 40% from the day Fields had taken over three years earlier until he got fired. Since then, shares have edged down further, closing at $10.64 on Thursday, the lowest since November 2012.

So for the new guy, there are a lot of things to review in his first 100 days. And on Thursday, at Ford’s “City of Tomorrow Symposium” in San Francisco – dedicated to “reflecting on the future of cities and urban mobility” – he reflected on the future of self-driving cars.

He’d been elevated from Ford’s “Smart Mobility” unit that works on autonomous vehicles. This move shows that AVs are a priority for Ford. But it is woefully behind Alphabet’s Waymo, according to data collected by the California DMV and reported in May for the year that ended on November 30, 2016.

The report shows how many total miles the AVs of various companies drove on public roads in California over the 12-month period without human intervention, and how many miles on average they could drive before human intervention occurred (the “disengagement”). Waymo’s vehicles drove on average 5,127 miles per disengagement; Ford’s vehicles, 196 miles:

  1. Waymo: 635,867 miles driven, 5,127 miles/disengagement
  2. GM/Cruise: 9,668 miles driven, 34 miles/disengagement
  3. Nissan: 4,099 miles driven, 28 miles/disengagement
  4. Bosch: 983 miles driven, 0.6 miles/disengagement
  5. Mercedes: 673 miles driven, 2 miles/disengagement
  6. BMW: 638 miles driven, 638 miles/disengagement
  7. Ford: 590 miles driven, 196 miles/disengagement
  8. Tesla: 550 miles driven, 3 miles/disengagement

So Hackett has his hands full. And in an interview at the symposium, he told the Wall Street Journal that Ford is rethinking how people will want to use self-driving technology.

Under Fields, the goal had been to have a fully autonomous vehicle without steering wheel or pedals on the market by 2021, to be used by ride-hailing companies – the driver-less Ubers. For Hackett, that schedule is still intact, but Ford is reviewing how to deploy the technology. And that might be different than envisioned before.

“The biggest leap is the nature of the human interpretation of using it,” he told The Journal. “If you think we’re going to take the [autonomous vehicle] and just replace the station wagon, I don’t believe that’s what’s going to happen. The AV will replace and do something that the station wagon can’t do – not just drive itself – but other things.”

He used the example of how computing technology took surprising turns, to where you can watch HBO on a smartphone.

“It’s about aligning the technology to what the market wants it to do – is it a new station wagon or is it an Uber vehicle?” he said. “We have work to do.”

AVs will redefine the car business. Alphabet won’t build cars, but it’s trying to control the technology that guides them. For automakers, this is scary competition from Silicon Valley. And so he said, since becoming CEO, he has been to Silicon Valley five times.

He has realized the importance of the infrastructure around self-driving vehicles, he told The Journal. This includes an operating system that allows everything to communicate and to coordinate:

“There’s a marriage of the evolution of the technology of the vehicle and the evolution of the system it works in,” he said. “In my 100-day review, I’m more convinced that the harmony of that is key to Ford.”

Alas, not only is Ford behind in testing self-driving vehicle technologies, it is also behind in connected-car services, such as updating vehicle software via wireless technologies, something that other automakers, including Tesla, have down pat.

In an interview with the Associated Press, also at the Symposium, Hackett was on the defensive over Ford’s acquisitions, such as the San Francisco startup Chariot which is driving folks around in San Francisco in 14-passenger vans and has since expanded its service to other cities. And to help out with artificial intelligence for AVs, Ford is investing $1 billion in a robotics startup, Argo AI.

Ford is undertaking these efforts while its dealers are getting restless with aging car and truck models – and they’re clamoring for a refreshed model lineup.

But Hackett said that these acquisitions had not taken money away from developing new models, and new stuff was in the pipeline. The amount Ford spends on the future is small compared to what it spends on car and truck development, he said. Successful CEOs have to prepare for the future and take care of the present.

People who live in cities now have more transport choices, including rideshare services, but personal car ownership will survive, he told the AP:

People already own vehicles and prefer to drive them knowing that they aren’t used 90% of the time, Hackett said. “I’m not ready to admit that Ford is facing a world where it’s losing a lot of personal buyers,” he said. “But I would admit that they’re going to have a lot more options.”

So it looks like the vision makes room for everyone in the foreseeable future: For people who insist on touching a steering wheel as they move down the road in a car they own, and for people who’d rather leave the driving to a robot they don’t own – and everyone in between. But as far as Corporate America is concerned, the impact of Silicon Valley might reshuffle the deck of the car business.

In Uber’s struggles to keep the cost of its drivers down but have them pick up customers in pleasing new vehicles, rather then in the old beaters, it has forgotten to do the basic car-business math. Now layoffs and massive losses loom in that division. Read…  Uber Gets Run Over by its Own Subprime Auto Leases

Enjoy reading WOLF STREET and want to support it? You can donate. I appreciate it immensely. Click on the beer and iced-tea mug to find out how:

Would you like to be notified via email when WOLF STREET publishes a new article? Sign up here.

  107 comments for “What Ford’s New Guy Said about the Future of Self-Driving Cars

  1. Truth Always says:

    Ford and even Tesla and others don’t stand a chance to Alphabet.

    The Waymo cars can drive much farther than anyone else without Human intervention so Alphabet (Waymo) can easily win the Autonomous Vehicle race.

    The real question is not Ford is why Tesla and Ubers continued to be over valued?

    Time to LONG Google/ Alphabet ?

    • Guido says:

      If alphabet could have won the race, it would have done so by now. It has been 14 years since an autonomous car drove in the DARPA competition. Even recently Alphabet was not so sure of making cars that can drive around by themselves. This is inspite of having tested cars day in and day out in mountain view. All that changed after Uber started its experiments of running the cars in Pittsburgh. But even then, the Uber cars carry a ton of computers in their trunk and need to be cooled.

      Computer vision is not at the stage a Literature major writes in Bloomberg — that is why they talk of LIDAR, laying out roads with multiple levels of sensors. If computer vision could figure out what’s on the road by simply looking at the image it gets, why then do these cars need sensors? The problem is that there are a lot of corner cases that these cars cannot cover. One intelligent reporter even wrote about how Google had to map out every street sign in Mountain View so that their cars could simply look up information from a database. That was why, the reporter explained, Google cars were seen only around Mountain View.

      As for Ford cars, they are testing them in San Ramon parking lots. Those lots are fenced off.

      I suspect this is a lot like what we used to joke about in software development. When a VP would show up inquiring if he could demo a running version to Wall Street analysts, the joke would be that he could use a PowerPoint as a back up with pictures on them of screenshots of said application, which would have crashed by then (requiring the ppt substitute). This would then have given the illusion that the application is alive and well.

      Self driving cars, like all engineering problems, have a lot of last mile issues to resolve, assuming everything else they claim is honest. Remember the story of a Mechanical Turk?

      • Truth Always says:

        Growing up in poor India and studying under kerosene lamps at night – I still get amazed to man’s technical ingenuity.

        I am close to 40 and would bet self driving cars would be everyday normal by no longer than 2025.

        As an extremely sceptical software engineer I drink no media cool aid but the way this software engineering has exploded in the last 10 year stupefies me. Maybe I am still a village boy.

        Yes Waymo has challenges but it’s clear as day how FAR ahead they are of everyone. APPLE is conspicuously missing. And Google / Alphabet is the company I most genuinely respect for their engineering. To paraphrase Trump ” It is going to be so beautiful that you are going to love it.”

        The only fear I have is that these technologies would contribute to sentient AI that would wipe humans. Elon Musk, Bill Gates and Stephen Hawkins are the only ones who seem to admit that AI IS Very dangerous. Everyone else is blissful ignorant or even hoping AI to take over.

        • alex in san jose says:

          Truth Always – you must have a very interesting perspective. We certainly had kerosene lamps, but they were only for when the power was out, because of storms or we couldn’t pay the bill. Generally when things got to the “couldn’t pay the bill” stage, we’d sold the lamps and just went to bed early.

          But a 1970s childhood was sure different than now. I remember many a time the electricity being out during the day and never noticing it. Families had one car, and it was a rite of passage to learn to ride a bike. I have the fondest memory of my father teaching me how to repair a punctured inner tube. It was taken for granted that you’d walk to school from age 5 or 6 on and in fact I never saw a school bus when I was in elementary school.

          Later, in high school, school buses were a thing. I think it’s because the school served a pretty spread out area, a long strip along the coast. I had the rite-of-passage of being on “Hubble’s bus”. Hubble was an older guy, I’m guessing what we called “hapa”, for half white and half Hawaiian, pretty dark skinned but with these blue eyes. He wore a red coverall and wore slippers that were too small for his big feet so he was practically driving barefoot. He was in a hurry to get out of there. So if you were on Hubble’s bus, you had to run like hell from class and jump a fence to get there before he took off. And if you were kept late or anything at all, well, you’d better have a dime to get the public bus home. Now when I hear talk about the Hubble space telescope I always wonder, “How did that guy ever get a telescope named after him?”

          Going to the library was a big deal, and viewing the 8mm educational films was great fun.

          We didn’t always have a phone because we couldn’t pay the bill. Once we even were on a party line, because it was cheaper. Generally there’d be one phone for every 3-4 houses and that was the emergency contact. There were phone numbers you could call for information like time or the weather. The time one was 543-3211. “At the tone, the time will be…”

          Now almost everyone effectively has a supercomputer in their pocket and their pets get better medical care than we kids did back then.

      • justme says:

        >>One intelligent reporter even wrote about how Google had to map out every street sign in Mountain View so that their cars could simply look up information from a database.

        Sounds about right. Much of the Waymo stuff probably is driving in known terrain. Which also explains why they can drive 5,127 miles/disengagement: It is mostly Mountain View. By the way, I have noticed MUCH less Waymo activity in MV since Waymo became an indepedendent company. The Waymo cars used to whizz by with their characteristic sound at all hours, but now activity has slowed to a trickle. Either they are saving money or they have moved to mere challenging terrain.

    • gary says:

      I’m sorry but this is all bulls**t. This new guy needs to come up with something to keep hope alive, that’s all this is.

      Watch HBO while in the car? So we get to watch commercials even when commuting? How wonderful.

    • Vespa P200E says:

      I think just a ploy to sell more cars…

      The drive behind self drive cars by Ford and other automakers may be that they see “carmegeddon” often referred by Wolf in making and need a new incentive for the folks to buy new cars.

      They see that many naive folks are locked up in their newer car payments for another 4 – 7 yrs to pay off. Hoopla over self drive cars is 1 way to move more volume…

      As for technology – Alphabet is probably way ahead of the game to sell the “system” of hardware and lest we forget tracking softwares for give away price to mine and peddle more data.

    • Shawn says:

      Winning a race is of little consequence. Show me some hard data of 1000 Families driving from the Bay Area to Disneyland and back on autonomous mode in its entirety over the course of a year. Show me that data and I’ll believe all the hype. And know his, as is the case with drug trails, a single fatality shuts down the study.

    • Mike says:

      It’s actually pretty simple, there are taxpayer subsidies to Tesla vehicles and pretty much all of Elon Musks’ businesses. When inefficient market players still exist due to decisions by politicians, and easy money from the Fed, you will never get true price discovery in any industry. Why do we need self driving cars anyway? So we can spend more time on our smartphones? As Americans, we have become a rentership society, and it does not bode well for future generations! Google aka Alphabet has been driving around photographing the public right of way for years, so they have the needed traffic data.

    • Mike says:

      A pox on all their houses, re self driving cars! Does anyone remember MS Windows? My computer word processor still freezes regularly. I am now looking for the WP X4 disk to reinstall it. My computer regularly becomes unavailable when it starts to update.

      Does anyone really want to drive a car that a glitch, or a hacker, or an order from some bankster can make plow into some concrete wall? Just try to prove a murder if it can be done from another country via the internet.

      I think enthusiasm for self driving cars will ebb when deadly accidents start to occur that result in death. I know enough about computers to know that major accidents are inevitable: my GPS tries to get me to drive off a cliff when giving directions to a relative’s house.

      Trucks are a greater concern: they can kill many more people. I think the core motive for the government funding of self driving cars, and other funding, comes from the sums paid to truckers and trucking firms. They are one of the few remaining sectors in our economy paid reasonable (not fair or generous) fees for work.

      The oligarchs cannot wait to remove this huge expense from their budget. The fact that many will have to die is just (to them) the cost of doing business. They will seek to evade paying for those social costs.

      • Kraig says:

        Then again stricter enforcement of speed limits would reduce deaths, yet so many manufacturers are against speed control system (machines won’t speed)

        • d says:

          “yet so many manufacturers are against speed control system”

          Speed and output limitations, along with GPS locators , get hacked.

          Even in big fleets.

          They increase the price of the product, cause problems, and the majority of the user base, dosent want them.

          To many rural and urban speed limits are put in place to protect poorly designed and maintained roads, and placate NIMBY voters not maintain road safety and good traffic management.

          Rather like merging stop lights on motorway/freeway on-ramps. They don’t solve the congestion issues, they simply move them from 1 place to another.

  2. BotfromTottenham says:

    Hi Wolf,

    Great article as usual. As a 70-ish guy with 50 years of radio comms and IT experience, including exposure to self-driving truck technology in the mining industry here in Oz, I am watching this whole ‘self-driving car’ meme with great interest and a fair bit of skepticism. If the ‘5127 miles per disengagement’ figure is true, I’m pretty impressed (and Waymo seems to be waaay ahead of its erstwhile competitors). My long experience of technological change (I started work fixing vacuum tube radios) indicates that major paradigm shifts like this take up to 50 years to become pervasive, for many reasons including raising the capital for developing and building the whole supply chain, including fuel (lots of new GWh), ultra-sophisticated maintenance skills, new regulatory regimes and other stuff like insurance contracts. I’m not a Luddite, but I think this paradigm shift is probably going to zig and zag for a few years before it settles down to being the ‘new normal’. In the short to medium term my money is on petrol/electric hybrids (which I notice Waymo uses as its test platform). And of course some countries are going to embrace it much faster than others – e.g. India only stopped building copies of a popular 1950s British car in 2014!

    • RD Blakeslee says:

      “…ultra-sophisticated maintenance skills…”


      Subaru dealer told me that there recent, computer invested cars are serviced by mechanics who just carry out computer-generated tasks.

  3. Keith says:

    Way further away then they think. Not in my lifetime.

    • Kent says:

      Agreed. Most folks don’t know that WayMo (hate that name) has all of Mountain View mapped out in high def 3D so it’s cars know exactly where they are in time and space. They know where every street sign, stop light, cross walk, and every other traffic regulation device is. Because it was entered by humans. And I’d updated every time there is a change.

      How expensive is it to scale this nationally? Not to mention other issues like big potholes, the fact that it will never be able to tell the difference between a shoe box and a concrete block in the road, detour signs put up the night before, the expensive array of sensors, not understanding what a persons facial expression is signaling, etc…

      635,000 miles sounds impressive until you realize it was done on the same few roads over and over. And someone possibly getting killed or injured every 5127 miles is not ok.

      This is still just an experiment and the experiment is still in its infancy. We will not know if it will ever work for quite awhile. Even if someone gets it to work reliably enough to replace a person, that doesn’t mean most folks will want to give up control themselves. Which means a smaller portion of the market will have to bear all of the extra costs. And that might kill the market for these things by itself.

      • jonlaughing says:

        I’ve given up control of my car already. No big deal. I simply take the bus and subway where I want to go. Walk to shopping and cafés and theatres. Or commute on my bike or on my feet. Why do I want my own car anymore? There are lots of rental agencies, limos and taxis. The best improvement would be an improvement in the street car. Cheap, fast and efficient. And many of them are driverless already. A driverless car is still a car – I just don’t want one anymore.

      • james wordsworth says:

        Then again who would have thought I could stroll down almost any street in the world on Google Maps a few years ago. There is up front cost but it can be done, and the big part only has to be done once.

      • Mike says:

        The annoying thing is that the billionaire, and millionaire oligarchs will cause the U.S. government to divert funding to map roads, etc. arguing that the public will benefit. I think that the benefit to themselves in profits, termination of many drivers, etc., is the main driver.

        We will then have more mass unemployment and loss of jobs of drivers of trucks and delivery vehicles. If robots can be developed to unpack and repack such vehicles, which is technology that is likely to be effective and not deadly sooner, that will mean the loss of ever more jobs.

  4. d says:

    This is a “Tech thing”.

    Phones systems were (and still are) a “Tech thing”

    India and various other third world nations leapfrogged from elites only on copper wire to everybody on mobile phone/internet very quickly.

    Ford was the only US automaker that didnt need a Govt bail out last time round.

    Is the only US automaker not running on indirect leftover clown troop handouts from the 08-16 incompetent US administration.

    In a technology dependent field the guy at the back is frequently teh guy holding some aces.

    Everbody keeps writing off ford ,and yet they are still here, by the work of their own two feet, alone.

  5. Nicko2 says:

    Smart cities in EU, UAE, China, Singapore, Japna ect… will have thousands of autonomous cars by 2020. The future is here. ….it just won’t be in the USA.

    • 2GeekRn0t2Geek says:


      I Think the 2020/2021 number is optimistic at best, and vaporware at worst.

      Driver less cars require will require:

      1. Zero Fault Tolerant software
      2. “Vision” (Currently Lidar)
      3. The ability to determine what that vision is seeing.
      (Also ties back to #1)
      4. The ability to interact with all passengers regardless of accent
      (Voice Recognition software)

      1. Requires lots of money and time.
      2. Still doesn’t work quite the way they would like it too.
      3. Still in it’s infancy, and if a DB of street mapping of landmarks and
      signs is what it takes for it to work, it will never get out of
      Urban centers.
      4. Voice recognition is still an iffy proposition. Just ask Siri or Alexa a
      Question without enunciating.

      This is a long term project, but the final result will be winner take all in regards to profits.

      My best guess is more like 2028-2035 for all 4 items to be completed, with the US and Asia arriving at approximately the same time. Because Intellectual property theft is big business in IT. Just look at Waymo vs Uber, and Epic vs TCS for 2 very recent examples of IP theft.

      • Bobby Dale says:

        Very good list, I would add the following, as it seems the engineers and management of these companies are unaware of practical considerations.
        1) Protection for the software companies and manufacturers from trial lawyers, who are waiting with bated breath.
        2) An emergency mode for escape in dangerous areas, think of Brightmoor in Detroit (or the finer neighborhoods in your locale). This brings up Federal intervention if the software companies “redline” certain areas.

        Anyone else, feel free to add to the list, or point out where I am mistaken with these two.

        • beadblonde says:

          AI and law don’t mix. AI is fine for something that can learn without consequence.

        • Bobby says:

          I’ll be laughing when a snow storm hits in Chicago, these cars won’t stand a chance even if it’s 2030..

        • Wolf Richter says:


          Nor do human drivers. At least AVs can “see” in a snowstorm. Humans go totally blind. Humans are totally worthless in a snowstorm.

      • TJ Martin says:

        Ahhh … bu the main thing they require which does not currently exist as well as being the single most expensive aspect of autonomous vehicles is ..

        Infrastructure .. with no … as in zero funds etc available to create it

        Add to that the inevitable litigation that will come with them .. an insurance industry reluctant to cover them … a general population unwilling to accept them .. along with the fact that autonomous or not a Car is a Car with all the inherit evils of the Car solving nothing in the long run being nothing more than technology for technology’s sake which history tells us is an ephemeral Rabbit Hole to fall down on the best of days … and I’ll place serious odds that like Flying Cars etc Autonomous cars will end up in the relegation bin of history

        What we need are solutions … not more complex problems adding to the problems we already have !

        And sadly for the automakers .. the only real answer is …. LESS .. as in less cars … less driving etc … L-E-S-S

        • T.J., not the real TJ says:

          Real TJ,

          I pretty much agree with everything you just said except “a general population unwilling to accept them”. Empirically speaking I have a lot of 6 figure women in my life who can’t wait to not have to drive themselves anymore.

          Also, only one of them that would consider giving up ownership.

        • david says:

          Everyone I have been around when the subject comes up says “no” to the idea of owning a self driving car. Myself included as I enjoy driving. They could have their place, especially with handicapped. Maybe cabs. Can a self driving vehicle back an 18 wheeler trailer into a tight loading dock? That’s a whole new set of problems. Wonder how much of a market will be there if it works?

  6. michael Engel says:

    Ford should design an emergency plan for smart mobility & agility, of itself to stay on it’s own feet, in the next recession.
    It will be harsh & long.
    Ford should avoid the self driving Ford back to $1.01 on Nov 2008(L).
    GM should definitely be in the same mode, cut the sleaze !

    • alex in san jose says:

      michael Engel – maybe this is why Ford rental bikes are everywhere in my area.

  7. RL says:

    Maybe they are out there but I haven’t seen a single study asking consumers if they even want autonomous cars. I’m old, a native Michigander (near Detroit) for 40 years, a West Coast guy for 20 years and I wouldn’t even consider a self-driver for my personal use. I live a very very low carbon footprint lifestyle but give up my car and my love for driving, selfishly, no way !!!!

    Turned in my 2014 Ford Fusion SE Eco-Boost lease car 2 months ago and I really liked it. After shopping hard for a month, I leased a 2017 Kia Optima LX for $100 per month less than a new Fusion SE could be leased for, couldn’t pass it up. Surprisingly, although it doesn’t quite have the creature comforts of the Fusion SE, it still stickered for $26,500 as did the Fusions I was shopping and I really,really like this Optima LX.

    Maybe Ford needs to get in the game. I was in the retail car biz for 25 years and know leasing in and out. Shopping a lease on the Kia was a breeze, shopping on the Fusion was a workout. I wanted to be local-dealer-loyal and lease American but for $100 per month I’ll drive 45 minutes and was treated super by the Kia store. In and out in 2 hours after an internet quote with no BS, refreshing for sure.

    • Wolf Richter says:

      Lots of people I asked would love self-driving cars (whoever owns them), assuming they work properly and are priced similar to normal cars (which won’t happen for a while … this is expensive small-production stuff for now).

      But you’re right, many consumers have no interest in them and may never buy them.

      But if only 10% of all consumers over 18 want them initially, that would be still be huge: 26 million in the US initially, and more as other people come around.

      • michael w Earussi says:

        Consumers won’t be buying them. Their future is in economical taxis.

        • Kent says:

          Completely disagree. You can get away with economical taxis in a highly urbanized environment. Folks in the suburbs/rural areas don’t want to wait 20 minutes for a taxi to show to go get the gallon of milk they forgot at the grocery store.

          Owning your own vehicle is vastly more convenient, especially when parking is never an issue. Same with autonomous vehicles: if I want to shave some drive time off and do 85 mph, I have that choice. I’m not sitting is some bubble car with no steering wheel doing 45 behind some big rig.

        • Dan Romig says:

          To add to Kent’s point, my family’s wheat genetics/seed company had our main production and experimental nursery plots at a farmstead six miles from the nearest store and town.

          It’s a different life living on a farm versus in a city. Our partner’s 13 year old daughter would drive a 3/4 ton pick-up truck to us in the fields with snacks at lunchtime, and she drove a combine at age 15. I reckon they’ll take a pass on self-driving vehicles.

      • TJ Martin says:

        Wolf – You’re missing an important statistic . Consumers between the ages of 18 and 35 don’t want cars .. period . Autonomous or not . Not only that consumers between the ages of 18 – 35 cannot afford a car .. any car . So that 10% is at best Pie in the Sky verging on delusional prognostication

        Sigh … to be honest .. every time I read about all this autonomous , EV etc garbage all of which creates more problems than they solve … the more appealing moving to CH becomes where trains trolleys and public transportation are the norm all but negating the need for a car .

        Book recommend ; ” The Technological Bluff ” by Jacques Ellul

        • Wolf Richter says:

          Whatch out that you don’t get run over by an EV or AV as you cross the street on a red light because you think these cars aren’t there /sarc

        • gary says:

          Actually, you point out a huge problem Wolf. I think electric engines are a lot quieter than ICE. You might not hear them coming. (I often hear an ICE car around the corner before I can even see it)

        • nick kelly says:

          “Between 18 and 35 don’t want car’

          ‘Between 18 and 35 can’t afford car’

          This is called making a virtue out of a necessity.
          ‘Look how cool and green I am, I don’t have a car’

          If they win, inherit, or even earn enough it’ll change.

          Of course if you live in a really dense urban environment, on the thirtieth floor of a pigeon coupe, you may genuinely not need a car, but I don’t know how people do that and stay sane.

  8. Halsey Taylor says:

    Built-in obsolescence seeks a return. Can you imagine how well Ford will support software updates for older models? Unlike hardware support, unlicensed 3rd parties can’t provide software updates because of DMCA. Not to mention, they can start charging annual software subscription fees. This is out there already, it’s Microsoft’s model.

    When it would like a revenue bump, Ford would be able to tank as many models as needed by stopping software support.

    • Wolf Richter says:

      Yes… see Microsoft. It stopped supporting Internet Explorer a few years ago. It has become so unstable that it’s practically useless.

      But I don’t think it’s going to be that easy for automakers – when a $50,000 vehicle that you own becomes “crashy” and useless because the automaker no longer supports the software, you’re going to see a consumer revolt against that automaker.

  9. mvojy says:

    Big business just can’t wait to destroy entire industries and professions with automation. The fully self-driving cars will eliminate car salesmen, mechanics, truck drivers, taxi drivers, limo drivers, bus drivers, etc. What other profession opens up for the millions of people that rely on these eliminated positions? Amazon and Fresh Direct will only hire so many.

    • michael w Earussi says:

      You forgot all the poor blacksmiths and buggy whip manufacturers who were put them out of business when the horseless carriage replaced the horse.

      • Steve Yerby says:

        The volume of blacksmiths and whip employees was about 1% of the auto industries now.

      • Kent says:

        Yes, the production of internal combustion engine was vastly more cost efficient than horses. Horses needed lots of land, feeding and care.

        And the government made ICE vehicles possible by paving roads. And cars paid for themselves because you could drive to a much better paying job. All of this created a vast industry that absorbed all of the buggy and whip manufacturers and teamsters.

        Automated vehicles will have none of the above advantages. It creates no new value. It might make taxis slightly less expensive, so people can spend slightly more on imports from China.

    • Nicko2 says:

      This is a good thing. The Earth’s population will increase by at least 2 billion people over the next thirty years. With the endless increase in urbanization and growth of megacities around the world (greater than 10 million each), automated vehicles will be essential. Of course, increase automation will open up new solutions and efficiencies (drone delivery, bicycle use, walk-ability, renewable energy, more mass transit, green space, ect…).

      As a millennial, I’m happy to declare I will never own a car. I can put those resources to much better use elsewhere.

  10. Paulo says:

    Yesterday, I hooked my boat trailer up to my 31 year old Toyota PU and headed for the boat launch. The road I traveled was un-marked. I crossed an industrial log sort to enter the gravel access road and suddenly a black bear leaped off the roadside in the front of my truck and disappeared into the estuary flats. I had to slam on the brakes. I reached the sort and prepped to launch in 4 wheel drive, inching down the bark and seaweed covered ramp until the trailer wheels were barely covered. After fishing for 1 hour I returned home as per, minus the bear….althoughy by this time I had to wait for a couple of logging trucks being unloaded. This required knowing how the log sort works and anticipating what would happen and my role that had to be played.

    Multiply individual challenges like this by tens of millions with a different spin, (the parking lot has a detour, whatever) and it is pretty clear that self-driving vehicles might work in a controlled pretend world where Stepford people commute on a freeway to their cubicle, but good luck the other 300 days of the year. :-)

    • Wolf Richter says:

      You’re very likely to keep driving your Toyota PU to tow your boat for many more years :-)

      Different transportation needs require different solutions. And that’s why there is such a large choice today, and that’s why there will be even more choices in the future, including AVs.

    • michael w Earussi says:

      Driverless cars will be primarily for the city. Small town and country folk will still own their pickups.

      • jonlaughing says:

        But any decent city will have buses, trains, subways, trolley cars, taxis, limos, sidewalks, bike lanes, and street cars. I don’t own a car in London or NY, and never will again. The car is an obsolete tool. But if I do want one the there are plenty of rental agencies.

        • Lee says:

          What is a ‘decent city’? We have all of the above and it is a mess.

          Melbourne has been voted the “world’s most liveable’ city again by the Economist for the 7th year in a row.

          The joke is that it is liveable except for those who live here.

          Transport in the city is a joke. Public transport is a disaster.

          The population growth here has been go fast and big that infrastructure can’t cope with the influx.

          An accident on the SE Freeway in the morning or afternoon peak will result in huge kilometer long traffic jams.

          Every week one of the train lines has a meltdown and commute times blow out.

          That 20 minute normal drive time in off peak blows out to 90 minutes or more. The 1 hour train ride……………turns into a nightmare with trains stopped for hours and no other way to get home or work. The longest it took me one time was 3 1/2 hours.

          Even in my little corner of Melbourne this bs is repeated every day during the morning and afternoon peaks. It is faster to walk to the grocery store than drive so we don’t drive or go shopping at those times.

          Want a parking spot at the local grocery store/shopping area around 10AM most weekdays? Drive around for 10 or 15 minutes to find one. We don’t shop at those times anymore. Need to go to the local post office?

          Again it is faster to walk rather than drive and try to find a parking spot during much of the day.

          Melbourne is out of control. Population is too big, but density is too small and to build the needed infrastructure is too late and too expensive.

  11. George McDuffee says:

    There appears to be two major types of vehicle contol programs: (1) Rule based programs created line by line by humans covering every possible contingency of incredible length and complexity [FWIW length and complexity promote instability and “bugs.]; and (2) Neural networks which “adapt,” but which humans have difficulty comprehending.

    The neural network adaptive/learning programs seem to be outperforming the rule based linear programs in AV operation in spite of considerably more money being invested [or spent] on the rule based approach.

    It is a scary thought that some [most?] of our vehicles being operated on the public roads will be controlled by programs which we do not and indeed cannot understand, but then again many of our drivers cannot explain how they drive except in the most general terms either.

    This technology of adaptive [learning] neural network control will be applicable to other areas such as aircraft and military vehicles such as tanks, artillery, and ships in addition to MotorT, and indeed “finance.”

    Anyone know which approach, possibly both, that Ford is using?

    • Wolf Richter says:

      The fact that Ford invested $1 billion in an AI startup tells me that they’re also pursuing your #2 approach (probably in addition to the #1 approach). I think all automakers know that AVs need at least some AI to function properly — hence the huge emphasis on AI, including the specialized chips needed for it.

    • Kent says:

      The rules based approach was, for the most part, abandoned in the ’90’s. The problem is that intelligence is really the ability to notice a pattern (a set of rules), abstract that pattern to a higher level, and apply it to circumstances that don’t obviously relate to the original pattern. The second two parts may be impossible for computers. At least no one has figured it out yet.

      So everyone is using the mass data approach. Google actually puts sensors on hundreds of human driven cars to understand their reactions in different situations. From that “learning” they can teach their autonomous cars what to do in the same situations. That concept is what everyone is doing in AI these days (at least on practical applications). And Google has great skills in this area simply due to their web search/ad business. Maybe better than anyone in the world.

      Here’s the problem with cars though: suppose 98.9% of the time, humans react identically in the same situation. And, by looking at the sensors, it is obvious why. But what about that extra 1.1%? Is it because people are doing the wrong thing? Or is it because their sensors aren’t detecting something that the human is?

      If it’s the first, AVs will improve driving safety. If it’s the second, AVs are going to get some nice folks killed or injured. But there is no way for Google to know without a lot of one on one personal interviews.

      • Guido says:

        Bingo! Remember how difficult it was to get insurance companies to get you a cheap insurance when you started driving? A new driver is 16 years old. The AI systems today have the IQ of 4 year olds.

        Now, one can always argue that IQ has nothing to do with driving and this is a false argument. Being able to interpolate a pattern from a sample dataset and correlate with existing knowledge base requires IQ. For otherwise, a 4 year old can design a self driving car once she is fed with all known knowledge. This trait may not be needed everyday but not having it when needed can be a matter of life and death.

      • Seems like AI for car driving is following AI for car engines. There is a parameter map for each sensor on your engine. If a non critical sensor craps out, the computer can figure out something is out of range, and it takes the engine off live sensor feedback and goes to pre stored maps.

        Testing will be a long way in for these self crashing cars. It will be interesting to see how it all gets done. Will these cars all have maps built in? will there be nightly updates for construction? What about morning wrecks…? Or evening wrecks…? How about northern roads vs. southern roads…? Up north they have no stop clover leaves. In the south there is a stop light at the end of each exit ramp, in the evenings traffic backs up onto the far right travel lane on the highway…. Exit ramps will be a single lane marked, but 2 or 3 lanes in practice as the traffic piles up. I think of stuff like this as “Driving slang” as far as road rules go. How will ai do that….? I read that FB shut down a AI experiment because the computers started to develop their own language that the programmers did not understand….. Sky-Net? Yah, On the fun side will our SkyNet cars have a free hand and will it develop it’s own hand insults for other Skynet cars that cut one another off…? Will they use our classic finger thing, or will it be a secret code……? More importantly, will auto factories be fully autonomous first? and will our AI cars have missiles built in without our knowing it so our cars fix traffic problems and leave an after-glow at the corner of first and I told ya so.

        It seems that Skynet cars will be a mixture of maps with helper notes, and some on the fly learning. Google is best positioned for this – using knowledge learned from all their advertising businesses.

        Infrastructure will be an issue. While in Atlanta the morning commute with “Ladder in the road on 285 East past I20” on the radio each morning I can’t imagine how these cars will handle it. I think it will be fun though….

        I look forward most though to seeing road rage between 2 Skynet cars….

        Should we even begin to think about hacking…..
        Slow networks………
        GPS sending cars with humans into lakes……. lots of them……
        GPS sending idiots into the desert where they make out their
        last will and video……… Idiots……. My Car tried to kill me by
        sending me into the lake……. Well stop cursing at your AI equipped
        car…… Oh-Yah Will they be called “Smart” cars….?

        Will there be Terminator cars !?!?!?


  12. Boatwright says:

    The fully autonomous self-driving car is a technoid fantasy that reminds me of the flying car in every driveway cover art we used to see in Popular Mechanics mag.

    Arguably, continuing advances in robotics and computer science have made self-driving cars conceivable. The big question is: Are they realizable?

    In a freeway type environment, with guiding elements built into the roadway, in good weather, with no pedestrians guided semi-autonomous vehicles may happen. This will require several decades of development and large investments in vehicle technology and road infrastructure.

    On city streets, parking lots, country roads, snow days, areas of driving that require human social interactions such as judging the likely actions of another driver or pedestrian through often subtle cues such as body language and eye contact — FORGET IT! As in other areas, complexity rules. The number of variable quickly overwhelms any current or planned computer system small and efficient enough to fit in a car.

    Well into the future, with quantum computing and the possible development of systems that will rival the human brain in complexity we may see “robots” that can think. Until then a few dozen tricked out cars with a trained safety driver behind the wheel buzzing around the streets in sunny California is a very poor indicator of the future success of investment in this area.

    I suggest this well written analysis by an expert in the field:


  13. Wilbur58 says:

    My thought has been that this technology is badly needed. Too many bad drivers on the road, exponentially compounded by the fact that many of these idiots are staring at their cell phones every 20 seconds. For safety reasons, we shouldn’t have people driving cars any more.

    I remember seeing someone recently comment, “A big part of the traffic problem is people needing to change lanes too much.” Well, some of us don’t feel like waiting behind people stopped for no reason other than their stupid phones. Or, some drivers are just too afraid and too slow. Those types used to stay on the right side of the highway. But no longer.

    I have a question for everyone. Have you seen Minority Report? I’ve always thought that the sort of “conveyer belt” system with pods that can attach and detach makes way more sense that free floating, self-driving cars. Is there anyone in the world even considering this model?

    Basically, streets become conveyer belts instead of open roads. Then, you get into a pod that’s just off the belt at an entry point. You program in your destination and then your pod moves onto the belt. At that point, your pod moves along the belt until the next turn where it would transfer to another belt. This continues until you reach your destination and then there’s an exit/entrance point for the pod.

    Isn’t this way safer? Is it really that much harder to engineer?

    • alex in san jose says:

      “Streets as conveyor belts” is an idea that goes back into the 1930s if not even earlier.

  14. beadblonde says:

    These manufacturers seem to know liability will be a trifle. I am puzzled. I’m glad I’m no longer at fault but I would expect lawyers to have a field day with automated vehicles.

  15. michael w Earussi says:

    I love the quote “I’m not ready to admit that Ford is facing a world where it’s losing a lot of personal buyers,”

    Car companies have seldom, if ever, embraced reality willingly. Ford is no different.

    The winner of this competition will be the company that combines a near perfect AI with a 1,000,000 mile maintenance free car. Ford’s expertise is in car manufacturing, not software. They should stick with making the body and just buy the software. Then they’ll succeed.

    • d says:

      “Car companies have seldom, if ever, embraced reality willingly. Ford is no different.”:

      FORD is the company that stood up at Detroit, just before the US auto bail out, and said “We have got to stop making what we want to make and start making what the consumer wants to buy.”

      They were laughed out of the room (Almost)

      GM still hasnt got that message.

  16. Steve Yerby says:

    Will driverless cars go with the flow of traffic or the speed limit causing more lane changes and accidents? Will the passengers insurance pay for damages or the manufacturer? Will a terrorist or lunatic hacker take over a truck at 70MPH on the highway?

    This removes the communication and cooperation between people. Supporters want to smoke weed in the back seat — it’s easier than learning a skill. Yes, the world isn’t perfect.

  17. Ambrose Bierce says:

    I don’t know how you are going to sell self driving cars to people who won’t use cruise control. What’s wrong with all this? The technology is built on the previous generations Buck Rogers fantasies and not organic models. The car is built on the model of the horse, horse power, etc. The electric car is disincarnate, and natural gas makes a lot more sense, but too many Star Wars movies have put the collective mindset in electric car neutral. The AV would be a great advantage for RURAL people who must drive hundreds of miles just to go the doctor. A lot more of us could leave the city if the gizmo nerds would take time to consider what the technology means.

  18. Guido says:

    There’s this thing known as loss function, i.e. Penalty for making a wrong decision. The fields you cited for neural networks allow you to make mistakes or lives are expected to be lost. With AV, you get very few chances. I don’t know how many Tesla drivers now watch movies and let the car drive by itself, but I suspect the number is close to zero (especially since they need the driver to man the vehicle at all times).

  19. Kent says:

    1. Waymo: 635,867 miles driven, 5,127 miles/disengagement
    2. GM/Cruise: 9,668 miles driven, 34 miles/disengagement

    The mile/disengagement is a really interesting number. Obviously, Waymo has solved a lot of problematic situations that GM hasn’t. But I’d be interested in seeing Waymo’s numbers over the years. It would be interesting to know if the curve is exponential, linear or logarithmic. That might give us some insight into when we could expect see a working product.

    If the goal is 200,000 miles (made up) between disengagements, is that 5 years away or 105?

    • Guido says:

      Unless the aforementioned flying cars show up in the mean time making the self driving car problem moot. :)

      • Ambrose Bierce says:

        at the end of the day why go anywhere, when Amazon will bring the world to your door?

    • The miles between disengagements is hard to judge because each car is not run one behind another on the same road facing the same decision at the same time. Each I imagine is tested in different cities etc. And what if GM’s is tested on a different road every day, while G’s runs the same road every day at least traffic time. What if they all don’t have a stored map and disengagements are caused by slow loading maps from a live network vs. stored on board map. If fords is really tested in parking lots then they have not even come up to navigating a road yet… 10 years behind….

      Remember the old windows 95 jokes “If Microsoft made cars” “If problem persists exit car close all windows and restart engine….”
      Scary to think that Microsoft could one day be driving a car….

  20. mean chicken says:

    The problem with humans driving is always the lose nut behind the wheel. This ticker looks like a bargain.

    • mean chicken says:

      Spelling mistakes are for your amusement. (loose)

    • Kent says:

      Actual story this morning in the “am I the loose (lose) nut” category.

      I’m driving into work down a two lane country (for Florida) road. I like to do 68 mph down this road which has a 55 mph speed limit. I come up on 3 cars, tightly packed, doing 50 – 53. I reckon maybe there’s an old gal up front.

      Nobody’s coming in the opposite lane so I decide to pass. Problem is, I start my pass on a double yellow line. Now, working for the gov, I know why and when they put down double yellow lines. So I know what the safety hazard is in the area. I check it out, I’m safe I go.

      Well the first guy I pass lays on his horn as I go by. He hates that I am passing all 3 cars, and maybe that I’m passing on the double yellow line. Don’t really know, but he’s mad. I giggle and keep going.

      After passing, I put the car on cruise control at 68. As I watch in my rear view mirror, the mad guy looks like he wants to pass the cars in front of him too. I’m thinking “go now brother, no cars are coming!”. But he’s timid. I know he wants to catch up with me and give me what I got coming. But he can’t do it.

      A minute later he makes his move. Bad timing. He’s got to on-coming cars. He looks like he’s gunning it (lots of smoke from behind his car) and the oncoming cars aren’t slowing down. Then both on-coming cars brake. Then the front car flies off onto the shoulder in a cloud of dust as my pursuer finally pulls back into the proper lane.

      I’m sure his heart’s pounding but he now has a chance to catch up. Unfortunately, some guy in pickup pulling a trailer with a little front-end loader pulls out just behind.

      To make a long story short, the guy tried following me in to work. He lost me when a school bus forced him to stop.

  21. Adam1 says:

    Some really good comments here, but I believe there is one huge factor that isn’t necessarily being considered – the relationship between cars, streets and pedestrians. Google a bunch of photos from the 1890’s or early 1900’s; streets are teaming with people intermingled with horses and carriages and even street cars. Pedestrians could easily predict the movements of large street cars because of the rails. Pedestrians could also readily navigate around horses and buggies because they moved at about the same speed as they did. By the 1920’s and ‘30s there were too many cars on the streets. Pedestrians were relegated to sidewalks. The risk of misjudging how a driver will behave when you step off the curb to enter/cross the street is just too high to do it haphazardly. However AV’s will likely change that risk and therefore the dynamic. What happens in a place like Manhattan when 30% or 50% of your cars are at a standstill because all the pedestrians, and there are a lot of them, know they will stop for them?

    • Wolf Richter says:

      Human drivers killed 38,000 people last year. Humans are notoriously terrible drivers. Humans go blind at night and in snowstorms and when it rains a lot. Humans get drunk or are on drugs or are tired or aggravated or distracted… Machines only have to be better drivers than average humans. And that’s pretty easy to do.

      • Guido says:

        “Machines only have to be better drivers than average humans. And that’s pretty easy to do.”

        Average human beings data set is huge as is the number of miles driven. To beat the average, the AVs need to do some serious driving under various conditions. With certain samples, it is easy to match. If the samples are hand picked (as is the case with Google mapping out the entire drive), beating the averages is even easier.

        Being better than the average at driving is the original problem AVs set out to solve.

        • George McDuffee says:

          One factor that seems to have been overlooked is the IoT [Internet of Things] and the car2car/car2base transmission of data. With a huge and growing AV base data set, and the advanced neural networks, rapid advances in autonomous vehicle control systems will occur.

          It is entirely possible that an autonomous vehicle control system will become “lost” on an obscure lane or pasture and will require manual control, but only once, as the data will be uploaded to AV central and integrated into the dataset. Every AV will teach every other AV, and every iteration of the control system will be an improvement over the last.

          While the creation of a neural net does appear to require considerable computer resources including talented systems analysts and ace TensorFlow* programmers, as many copies of the evolving neural net as desired can be duplicated by relatively unskilled [i. e. low cost] workers, which will run on much less powerful computers in the vehicles.

          * http://playground.tensorflow.org/

        • Guido says:

          Terrific idea.

          I am guessing they are still trying to figure out what works and what doesn’t. In fact, what you suggest is a good test of how ready the cars are for prime time. If they have figured out enough to incorporate common knowledge into servers and share it across their own cars, we know that they have hit their stride.

  22. Petunia says:

    A couple of months ago I saw one of the driverless cars on a well traveled street in my town, southern flyover country. I noticed it at the light because it had some tripod thingy sticking up on top, then I noticed it was driverless. I was so surprised I didn’t look at the markings, but there were some on the car.

    The car left me somewhat scared because one of the other streets I frequently travel on has been under repair for the entire time I’ve lived here. They are frequently changing the lane patterns to divert the traffic.
    I can already see, you don’t want to be on that street with that car in the mix.

  23. Kreditanstalt says:

    We don’t all live in large, modern U.S. urban centres. We don’t all have breadwinner jobs. We’re not even Americans. And few – VERY few – of us have ever used a smartphone to take a taxi…we don’t use taxis and some can’t afford cell phone service at all. (And out in the bush there isn’t any)

    We do live in places where the drive to Bamfield includes a 85km gravel road, potholes and all…and will I get a self-driving pickup to load firewood? Nor am I personally sure I want to be completely dependent on a centralized electricity grid to start my car…

    Not going to happen. Tech-worshipping candy rainbow-farting unicorns.

  24. Kf6vci says:

    What’s the big benefit to society and car owners again? Less many accidents? No more cars being used by terrorists? Governments’ ability to “brick cars”, trumping Orwell’s 1984?

    War on cash. No more bank runs. No surprise that China is at the forefront of that movement. “Chipping people”? Or make that “phone verification” for all credit card transactions. Voila, total control.

    And that’s gonna be a wonderful new worls?

    Murksel in Germany calls them “Zero Emissions Vehicles”. Yeah, no more laws of physics here. Have faith. Those 3 million jobs in Germany will be replaced by illegal immigrants paying taxes one day. (Sarcasm off)

  25. JR says:

    Miles per “disengagement”. LOL. From the news reports I see – truck drivers are regularly disengaged by their smart phone – to the point where they don’t notice that traffic has stopped in front of their truck. They usually will consume a car or two while reducing their kinetic energy to zero. Entropy in action. And it helps keep the replacement auto sales going.

  26. ft says:

    The basic trouble with self-driving cars is the same (only more severe) as the trouble with GPS. With GPS, drivers enter a destination, follow the instructions they are given, and arrive at the correct place almost every time. Wonderful except for one thing: GPS allows you to get anywhere you want to go without ever knowing where you are. I see people do this all the time. The geographer in me is appalled! We might as well just skip self-driving cars, don a set of virtual reality goggles, and never actually leave home at all.

  27. raxadian says:

    Self driving is for cars what “3D” is for movies only with way way way more deaths. Did yoy know the earliest 3D system for movies did not need no stupid glasses? The first 3D glasses for movies were invented because filming with two cams and then merging it in a single film was expensive.

  28. Hiho says:

    “Ford is undertaking these efforts while its dealers are getting restless with aging car and truck models – and they’re clamoring for a refreshed model lineup.”

    Ford had better take care of its core business rather than going after a techonology that will not go anywhere. But well, I guess that this is not so cool.. and we want the media to hype stock price up, don’t we?

  29. Dan says:

    Has anyone considered that “Self Driving Cars” = nothing more than a unicorn scam designed to create a whole new host of virtual businesses that feed off the ‘idea’ that one day they will be profitable?

    The downsides of self driving cars that will make them intolerable for anyone human.

    1. Your insurance company will know where you are 24/7. Hence the police will follow you 24/7.

    2. It’s likely your employer will follow you 24/7, and require you to work in your car. Didn’t think you were going to catch up on zzz?

    3. No more affairs and google or Ford or whoever will use the car cameras to photograph you and your neighbors.

    4. People will be killed left and right as these things careen off the roads in bad weather.

    5. Car maintenance will be a nightmare and you’ll have to keep your car in primo condition to prevent lawsuits if you get in an accident. No more beaters, new cars only for the next generation.

    6. local municipalities will go bankrupt trying to keep their roads ‘perfect’ for these machines.

    7. Because of the above expenses its likely your employer will own these things and send a car to pick you up in the morning (welcome to the new serfdom).

    This is the biggest unicorn con yet!

    • Wolf Richter says:

      This is hilarious. You just made my day. Your points are really funny. I just couldn’t resist. So item by item …

      1-3: Already happening in the current vintage of human-driven cars. They have cameras and they’re connected to the internet, and this has been around for a long time. For example, GM’s OnStar which started its service a decade or two ago.

      4: Already happening by human drivers who killed 38,000 people last year in traffic accidents. Humans are blind in bad weather and after dark. Sensing technologies are not.

      5: these will be electric cars, and electric cars are very simple compared to ICE cars and require little maintenance (no ICE engine, no oil change, no filters, no transmission or transaxle, no emission control systems, no exhaust, no coolant or radiator, no belts and hoses…) . Check out what’s under the hood of a Tesla… storage space for your groceries and luggage instead of an engine.

      6: Local municipalities are already going bankrupt… see Detroit and soon Chicago, plus the four that filed for bankruptcy in California, plus Jefferson County (Alabama), plus… AVs have the same problems human-driven vehicles have with bad roads.

      7: Employer will likely send a car to pick you up…. Currently only CEOs of the biggest companies have that privilege. Rank and file employees have to commute on their own. They would love it if they don’t have to pay for the commute and fight it out themselves on the congested streets and highways. That would be one of the biggest benefits a company could offer. Forget the Google buses… you’d get picked up at the door and dropped off at the entrance of the building where you work. On the trip, you could sleep or work or just stare out the window wondering who all the poor fools are who still have to drive to get to work. Where can I sign up?

    • Guido says:

      You are assuming that the people who are spied upon won’t react. For example, in East Germany the Stasi had so many informants that people were not sure who amongst them was a spy. So they came up with this weird way of communicating — they’d use sarcasm, reverse jokes, substitute nouns with some other nouns, and inside knowledge and communicated the way lovers would in public place. If you did not know the context, you’d not know what they are talking about.

      Once people react, they’ll probably devise a myriad ways of leaving so many false trails that mining that data for patterns will be probably as difficult as it is today, only with more data.

      Let me give you an example. Back in 90s when truck owners installed gps trackers on their trucks so that the drivers only drove for a fixed number of hours instead of all nighters, the drivers were initially baffled. Then they figured out that putting a metal bucket on the antenna stopped the data collection. They didn’t know the physics, but the solution was correct.

      • Nicko2 says:

        People will not react, they will be entertained and pleasantly distracted. ;)

  30. Mizuki Park says:

    Driverless cars will be primarily for the city. Small town and country folk will still own their pickups.

  31. ian says:

    ”Under Fields, the goal had been to have a fully autonomous vehicle without steering wheel or pedals on the market by 2021, to be used by ride-hailing companies – the driver-less Ubers. ”

    I have said this countless times, where are the regulators? How is it possible for a CEO to make such a goal? No wheel or pedals? We are all regulated to within an inch of our lives but just like that a CEO can decide that a car has no steering wheel?

    I just don’t get it. I also don’t get where they think the demand will be, who wants this? How many people are interested barreling down the highway with their feet on the dash while a computer does the rest? At night? Hell no.

    Then we have the other nugget, millennials are not interested in owning stuff, we will share everything. Yeah right, maybe that’s because they have no jobs, are living with their parents, get paid minimum if they have a job. Just maybe when they are older and have finally clawed their way up they will want possessions just like the countless generations that have gone before. People LOVE their cars, this is going to end in a matter of years? Bullshit. And I can just see me deciding to go on a caravan holiday in France, so what do I do, phone up uber and say can I have a driverless car for 2 months to tow my van down to France. Utter nonsense.

    • George McDuffee says:

      You have not spent enough time inside corporations and other large organizations.

      Many of the CEOs and executive vice-precedents, what ever their actual title, are self absorbed sociopaths/egomaniacs with disillusions of “grandeur” to which the laws of physics, the principles of economics, and GAAP do not apply.

    • Wolf Richter says:

      “…where are the regulators?”

      Regulators at the federal and state level are all over this, and the US Congress is working on laws to accommodate self-driving cars.

      Also, insurance companies are all over this, trying to figure out liability issues, crunching the numbers on actuarial issues, so that they can offer coverage when these cars are ready.

      • ian says:

        Well Wolf, maybe you think it is great for us all to be treated as guinea pigs on the roads that we pay for just for the benefit of these companies but i do not. It also shows how little these regulators, which you are so clearly impressed with, think of the common man. This driverless lark when it really gets going will cost millions of good paying jobs for people with low qualifications and no they will not all become geeks doing the programming. There is absolutely no benefit to the government in this, things can continue just fine without it. They could quite easily delay this almost indefinitely by banning them testing on public roads but they don’t give one shit for the people just their paymaster lobbyists.

  32. Kreditanstalt says:

    There seem to be about equal measures of rainbow tech-worshippers and doubting Luddites here…

  33. PRice says:

    I wanted to congratulate Lee Aug 18, 2017 at 3:35 pm for telling the truth about Melbourne, but his post has no Reply button. What gives?

    • Wolf Richter says:

      The system allows for 4 “nested” levels of replies (so one comment, a reply to the comment (=reply 1), a reply to reply 1 (=reply 2), and a reply to reply 2 (=reply 3).

      With each level of reply, the column gets narrower, which gets to be a real problem on smartphone screens. Hence the limit. You can reply to a higher level comment (with a reply button) in the same thread and address it to “Lee.” If he follows the comment thread, he’ll see it.

  34. Roman T. says:

    quote: “He used the example of how computing technology took surprising turns, to where you can watch HBO on a smartphone.”

    He considers watching HBO on a smartphone a surprising turn. That sums up his lack of understanding of technology, its growth speed right now and also opens the door for him to be in any other role in the company or just up and out.

    Sure personal car ownership will survive but so, so many will choose the cheaper and more convenient alternative. Especially the ones that cannot afford a car now.

    No parking problems, tickets,finding a spot, paying for parking, massive,expensive parking structures with stairwells filled with feces, no fender benders to worry about, no speeding or DUIs.

    Also cops cannot harass as they can do now. Can’t pull you over for a bad tail light and honestly I have no idea how they will pull one over. (There must be some features around that.)

    Lot of upside to the dronemobiles. Some dark downsides too. If a bad actor takes control of the vehicle you’re in and drives off a cliff for example. This is a huge issue I’ve not seen mentioned.

Comments are closed.