Catch up on stories from the past week (and beyond) at the Slashdot story archive

 


Forgot your password?
Close
typodupeerror
×
Transportation Businesses Robotics The Almighty Buck

Self-Driving Cars Likely Won't Steal Your Job (Until 2040) (wired.com) 129

Posted by msmash from the closer-look dept.
The self-driving robots are coming to transform your job. Kind of. Also, very slowly. From a report: That's the not-quite-exclamatory upshot of a new report from the Washington, DC-based Securing America's Future Energy. The group advocates for a countrywide pivot away from oil dependency, one it hopes will be aided by the speedy adoption of electric, self-driving vehicles. So it commissioned a wide-ranging study by a phalanx of labor economists to discover how that could happen, and whether America might transform into a Mad Max-like desert hell along the way. The news, mostly, is good. For one, self-driving vehicles probably won't wreck the labor market to the point where gig economy workers are hired out as mobile blood bags.

In fact, they'll eventually feed the economy, accruing an estimated $800 billion in annual benefits by 2050, a number mostly in line with previous researchers' projections. Two, robo-cars won't disappear the jobs all at once. "We have a labor market characterized by churning -- continual job creation and destruction," says Erica Groshen, a visiting labor economist at Cornell University and former Commissioner of Labor Statistics, who worked on the report. "The challenge is to make the transition as smooth as possible."
This discussion has been archived. No new comments can be posted.

Self-Driving Cars Likely Won't Steal Your Job (Until 2040) More

Self-Driving Cars Likely Won't Steal Your Job (Until 2040)

Comments Filter:
  • sure sounds more reasonable than what I have been hearing.

    Just my 2 cents ;)

    • Re: (Score:2, Insightful)

      by justthinkit ( 954982 )

      Truck driver is the number one job in most states. Tesla already has shown an electric truck, that is probably already self-drivable.

      This is going to hit a lot earlier than 2040.

      • Re: (Score:1)

        by Anonymous Coward

        You’re a total retard if you think Tesla is anywhere near having real self driving [jalopnik.com]. Tesla's auto pilot is the self driving equivalent of a carnival side show.

        • Re:That time table (Score:5, Insightful)

          by LostMyBeaver ( 1226054 ) on Thursday June 14, 2018 @10:58PM (#56787802)
          Actually, yes, I believe Tesla is much closer than 22 years away from having a self-driving truck. Far more than a self-driving car at least.

          Trucks drive a lot differently than cars. Most truck driving is highway driving. A large part of truck driving... especially the type that is is associated with jobs counts is logistical from business park to business park. It is entirely possible and likely that companies that are moving stuff from docks to warehouse or warehouse to warehouse can very easily be made self-driving friendly.

          Also, with the exception of managing traffic diversions due to construction (which I haven't seen yet on self driving vehicles), trucks can make the majority of their transit in the a single lane on the highway.

          Also, it could be possible for a business to arise for "last mile operators" who are vehicle operators that are responsible for navigating populated areas in trucks. As such, they would assist the truck from the loading dock to the highway and then be picked up by a shuttle bus. Then they could be delivered by a shuttle bus to the highway and assist trucks the last mile to the unloading docks.

          I am very much under the belief that we will accomplish self-driving trucks long before we achieve self-driving vehicles that could navigate my neighborhood.

          • Trucks drive a lot differently than cars. Most truck driving is highway driving.

            Even so, autonomous trucks will still require all the same "abilities" that autonomous cars will need. And then some more. They will still have to navigate normal roads (the "final mile" problem) and deal with the vagaries of other road users.

            So if the critical path to AV deployment is the technology, the limiting factor is still development of the systems needed.

            I can see that once AVs gain a foothold on roads, the pressure to go fully AV will be very strong and very rapid. Possibly to the point where

            • They will still have to navigate normal roads

              Huh? Most trucks aren't allowed on normal roads. They leave the warehouse and drive on the highway where the exit will take them right to the next warehouse. That's kind of why warehouses aren't actually positioned in city centres in the first place.

              and deal with the vagaries of other road users.

              See above: The other road users will be mostly trucks. The last mile problem for trucks is typically more accurately described as the last 200 yards. GP's point is that it's order of magnitudes easier to do this with a truck and they're right.

              • They will still have to navigate normal roads

                Huh? Most trucks aren't allowed on normal roads.

                You'll need to provide a link for that. In every state I've ever lived in, trucks can go anywhere that isn't specifically marked "No Trucks".

                • So you've never noticed the bridges near you with height/weight limits? Height/weight limits which preclude a very many large trucks?

                  All bridges have them, and a lot of them won't safely support a semi. Drive off any major road sometime, and you'll see plenty of them. They're not marked "no trucks", but they are definitely "no trucks" given the height/weight limits.

              • Re: (Score:2)

                by pnutjam ( 523990 )
                I've worked in plenty of warehouses that took deliveries from semi's in city centers.

                • Congratulations. I drove a car on a beach therefore all cars should need to be able to drive on a beach to be roadworthy right?

              • Trucks need to drive on normal roads to get to the highways, and they need to deliver to stores which are often in urban areas.

                • Trucks need to drive on normal roads to get to the highways

                  No, some trucks need to, and re-read my post because you seem to have missed its point.

                  and they need to deliver to stores which are often in urban areas

                  No, the kind of trucks we are talking about are not allowed to drive in urban areas.

          • Take any of those long cross-country trips you're talking about with a lot of highway miles and you're probably going to come across at least one place where the road is under construction, probably several. Often enough, that's out in the boonies where there isn't a continuous feeder, so there is a detour through the town square - exactly the kinds of situations automated systems can't deal with are going to happen. Drive across several states and very likely part of the journey is going to be through a sn

          • If you're going to need "last mile operators" why not just build a half decent train (railroad) network and use that? Self driving trains make a lot more sense than self driving trucks. Barring freak accidents, they don't have to be able to cope with other cars, cyclists, pedestrians, etc.

          • Even if they aren't all self-driving there's already been demonstrated examples where trucks have gotten from A to B without human intervention thanks to platooning. Just because a human driver still needs to be involved in one vehicle doesn't mean several others can't be autonomous behind them.

        • Re: (Score:3)

          by MrL0G1C ( 867445 )

          I wouldn't disagree with that but Waymo appear to be years ahead of the competition, 5600 miles on average now before the driver needs to intervene, 100 times further than uber vehicles.

      • Re: (Score:1)

        by zifn4b ( 1040588 )
        Why is this such a problem? Commercial truck driving has to be one of the most boring, lonely jobs available. Introducing self driving commercial vehicles not only eliminates this waste of use of one's existence but also makes commercial transportation more reliable. It's remarkable to me that when John Schumpeter's Creative Destruction happens again and again in Capitalism we have the same reaction as we did to the Ice Industry and the Tailor/Seamstress industry that became automated. The only concern
        • 1) You don't get to decide what work other people should find to be a waste of one's existence. Plenty of people would rather not have to deal with coworkers or customers, and like time on the road.
          2) As much as I'd like to live in a word where automation makes it possible for everyone to live without worrying about where the money for food, rent, utilities, etc comes from, we're not on that path. The people whose jobs get automated don't get taken care of - they get screwed. And I don't see any reason

          • Re: (Score:2)

            by zifn4b ( 1040588 )

            As much as I'd like to live in a word where automation makes it possible for everyone to live without worrying about where the money for food, rent, utilities, etc comes from, we're not on that path.

            We won't EVER be "on that path" while you consider your self interest more important than technological progress and innovation. It's known as Tragedy of the Commons [wikipedia.org]. This site's community has gone to hell. It's no longer intellectuals and insightful people. It's infested with toxic luddites. You'll reap what you sow tho. Mod me into oblivion but you will reap what you sow and you know what? People like me that said "I told you so" will let you fall into the abyss without extending a helping hand.

            • You're a damned fool if you value technological "progress" and "innovation" more than your own ability to provide for yourself and your family.

              Also, you might want to actually read that article you linked to. This is not a case of a public good being destroyed by people acting in their own best interest. It is a case of people being skeptical that those who are pushing advances have considered the consequences of their actions, and that those who hold all the resources will be at all interested in prov

    • It sounds pretty stupid to me; why would slashdot readers be presumed to work as taxi or delivery drivers?

      I thought most of the people here did tech work.

      Self-driving cars will steal my job never; why would a software-writing AI be given a street-legal automobile as a freakin' case?! I'm thinking the AI that steals my job will fit in a standard 19" form factor server rack.

      • It sounds pretty stupid to me; why would slashdot readers be presumed to work as taxi or delivery drivers?

        Because according to the Uber groupies here, being part of the exciting new gig economy by driving strangers around at 4 in the morning after you've already worked a 12 hour day coding is just...fun. Plus it's striking a blow for freedom: you don't have Uber drivers in Soviet Russia/North Korea after all!

    • Re: (Score:2)

      by bigpat ( 158134 )

      sure sounds more reasonable than what I have been hearing.

      Sure, it usually takes 15 to 20 years for even a wildly successful technology to permeate an economy from the first commercial implementations... Blackberry (1999) and Palm both had successful smartphones in the early aughts, but the technology really didn't take off until 2008-2012 time frame. And you could argue the less powerful Palm PDAs of the 1990s and were the pioneers in that commercial space and smartphones today are just the evolution of those devices.

      The twenty teens for smart cars are like the

  • the legal framework self driving cars will take time as well the uber death likely slowed things down.

    • Re:the legal framework self driving cars will take (Score:4, Interesting)

      by fluffernutter ( 1411889 ) on Thursday June 14, 2018 @07:54PM (#56787132)
      Uber or not, it's going to take a long time to work out who takes responsibility when something bad happens and there is no steering wheel. You can't hold a person responsible for an accident when they are always just a passenger and never a driver. It's their property and they should cover it against theft and damage by vandalism and other unfortunate events. But someone else has decided how that vehicle will behave in every possible situation, so a person should not insure for what it does.

      • Uber or not, it's going to take a long time to work out who takes responsibility when something bad happens and there is no steering wheel.

        This is already "worked out". Manufacturers are responsible for their products. When brakes failed on Toyotas, it was Toyota's problem. When Tesla Autopilot crashed into a truck the same color as the sky, it was Tesla's problem.

        • No manufacturer seems prepared to treat software failure and mechanical failure as the same thing; even though with the on vent of self-driving, we are resting most of the passenger's safety on the ability for software to correctly interpret any situation. Feel free to cite references if you find any evidence to the contrary. Furthermore, Tesla has carefully structured their liability such that they can always blame the driver. Every accident by autopilot has been followed by comments to the effect of, "

  • Lot of potential (Score:5, Interesting)

    by pubwvj ( 1045960 ) on Thursday June 14, 2018 @07:38PM (#56787058)

    There is a lot of potential and not just on the road. Out here on the farm and in the forests self-driving tractors, skidders, buncher grabbers, conveyers, wagons and delivery vehicles have a lot of potential. They are levers that amplify us. Just as it is easier to hammer in a nail with a hammer than your hand it is easier to move round bales of hay with a tractor than by hand. Self driving tractors would let me instruct them to put out hay to our pastured pigs in the winter (hay replaces fresh pasture) rather than my having to drive the tractor. Then I am freed up for other tasks.

    • It makes a lot more sense to develop self driving on equipment like that, with few obstructions and on private property, then to develop it on public roads.

      • One of the companies in the self driving car space I follow news from is Voyage [voyage.auto]. One of the areas they have a test vehicle in right now is a retirement community.

        This is a really great use I think of self-driving car tech, because it gives residents who may have trouble driving as they age a convenient and safe taxi they can use 24x7.

        These are public roads but are a good first step as retirement community roads are more laid back than most neighborhoods...

        There are a ton of older drivers that self-driving

        • I can understand that. They can move slower too, so would be less likely to miss stuff. It's possible to map a small area accurately, and you control what happens so if there is construction, lets say, the contractors that do the construction can be relied to update the map properly or use markers that are sure to be understood by the cars. You can make sure all the signage is in good shape too, and people are less likely to mess with them.
    • I have a client that is a harvesting company they are starting now in the south and will work their way to the Dakotas. Partly Self driving Harvesters are already being tested in the fields. But they still have an operator in the cab. Since I am not with the crews I am not sure of the extent.

      Just my 2 cents ;)

    • Just as it is easier to hammer in a nail with a hammer than your hand

      Wuss.

  • The churning labor market idea is obsolete (Score:5, Insightful)

    by presidenteloco ( 659168 ) on Thursday June 14, 2018 @07:49PM (#56787114)

    "We have a labor market characterized by churning -- continual job creation and destruction,"

    That is OLD SCHOOL THINKING. No longer applicable.

    That will no longer be true once AI and automated systems capabilities generally get better than the corresponding human ability.
    Example: There's a technology that is better now at detecting certain types of tumours in images than radiologists.

    We have to change our analysis of future job prospects, and not just rely on "something else will come up for people to do."

    There will be a cross-over point for each type of job when automated system will be better at it and more cost effective than a person.
    That will start happening to more and more job categories (or at least their most important tasks) faster and faster, as AI and automation continue their rapid advancement in capability.

    Automation and AI are improving fast.
    People are not.
    Get used to it.

    • The truth is, the days of getting a good job and sitting back, waiting for retirement are gone. Everyone must take personal responsibility for their own education, continuous professional/trades planning and training. Those that do not or are not able, will suffer for it. I know it is sad, but that is the truth everyone working in our current and our future economy face.

      Just my 2 cents ;)

      • How come in the last few years, whenever somebody says the words "personal responsibility", he always uses it to justify horrible things happening to people?

      • responsibility for their own education 50K a year + the cert treadmill = big loans with small hope of paying them back.

    • There's a technology that is better now at detecting certain types of tumours in images than radiologists.

      In one, carefully curated study. When doctors start actually being replaced by that technology, we can say it's better, but for now all we have is a preliminary study. Here is just one of many reasons it matters [xkcd.com].

    • Yep, It's the fundamental mistake people who use the lamplighter or buggy whip analogy make. AI isn't coming for one job. It's coming for all jobs.

      Including making bad analogies.

      • I concur.

        And an often missed issue is even if someone is correct in saying, "AI won't replace my job before I retire.", they are ignoring that entire industries are going to be replaced by AI. If their job is tied to one of those, it may not be replaced by AI, but rather destroyed by it. And even if AI doesn't destroy your industry, it might shake it well enough that a lot of companies go out of business, and that could include the one you work for in your "safe from AI" job.

        I analyze and design a lot of sy

    • I think the "churning" is supposed to be from one job type to another. As they say, if you're a buggy whip maker, you'll just have to train to be a software developer. Or something.

    • Re: (Score:2)

      by JBMcB ( 73720 )

      That is OLD SCHOOL THINKING.

      That may be true this time, however, the same thing has been said for every previous technological advance. THIS time there will be nothing else for the field hands to do. THIS time auto workers will have nowhere else to go. THIS time secretaries won't have any other job.

      If you could predict what the next major employment sector will be, you would be a very rich person.

      • That may be true this time, however, the same thing has been said for every previous technological advance.

        You're doing inference from stupid. It's like wisdom of the crowds in reverse: round up all the people who've been gloriously wrong (over and over again) into a small pen, and then go opposite George.

        News flash: you can't squeeze a correct prediction out of a teapot of stupid people.

        The fall of Rome was predicted many times. These predictions were wrong every time—until it actually happened (onl

    • I would argue that the churning labor market idea isn't just obsolete, it's flawed to begin with, hiding what really happened. What they fail to mention is that the people whose jobs were destroyed in previous technological revolutions struggled mightily to gain one of those newly created jobs. Much like how today you can't take an old mill worker and turn them into a AI coder, they couldn't simply take a serf, farmer or craftsman and turn them into a skilled factory worker. In the end, those people were
    • Labor churn is currently happening, and it will for quite some time. Yes, at some point, computer and robotic systems will probably get good enough to render most human labor obsolete and surpass human mental activity at the highest levels of strategy and creativity. The uncertain question is when. Next decade? No way. In 50 years? Maaayyybeeeee but I'm very skeptical. I'm guessing more like centuries. And that's only if we don't experience planet-wide wars or major civilization setbacks.

      In other words
  • The hype has overcome the reality of what can be delivered. Self-driving cars don't work yet. They don't.

    .
    I challenge anyone who says they do work to cross the street on a dark, windy, snowy night while self-driving cars are coming at them.

    Then there is the liability aspect. Who is responsible when a self-driving car goes awry?

    Toast.

  • Retirees will outnumber workers in 2030. More retirees mean fewer jobs.

  • That's only 22 years away (Score:3)

    by rsilvergun ( 571051 ) on Thursday June 14, 2018 @08:02PM (#56787178)
    most of the people reading this will still be alive then. For the older set who actually have the time an inclination to vote now's the time to do something about it. If the younger lot can't work your retirement's going to collapse with the rest of the economy. If you let that happen then you won't even be left with dog food.
    • What are we voting for? Saving the buggy whip makers?

    • Re: (Score:1)

      by Anonymous Coward

      It is sad that people are so weak in imagination or indoctrinated into the existing system that they can't imagine how we might manage a future in which fewer, or eventually no, people have to work. That is what I dream of for my children.

      It will come. America seems set on resisting it to the end. Other countries will embrace it in our place. We will wall ourselves off and sabotage it for as long as possible because it doesn't provide as well for allowing an elite few to maintain their exclusive position.

      It

      • Every time someone talks about a future where people can afford to live and not work, they get called down as a communist. I think many people can envision a world like that. Heck, Star Trek was about a world like that. But what is difficult to envision is how we transition to an economy like that because it requires the people with all the power to give up what they have so others can live. People get called communist, and reminded communism never works. People imagine it all the time, and then get os

      • It is sad that people are so weak in imagination or indoctrinated into the existing system that they can't imagine how we might manage a future in which fewer, or eventually no, people have to work.

        I don't think anyone has a problem imagining that world. I think they have a problem imagining how we could possibly get to such a world. The big problem is that the incentive for implementing automation is to not have to spend the money on workers. If the owners of the means of production aren't willing to pay people to do work, why on earth would you expect that they'd be willing to pay people to NOT do work?

    • Re: (Score:2)

      by dissy ( 172727 )

      If you let that happen then you won't even be left with dog food.

      Actually, retraining the displaced market and economic analysts into a future job of becoming dog food was our plan all along!

  • Semi intelligent cars will not be fully autonomous for a long time. But they can be driven remotely. One operator in Mexico could monitor several trucks at the same time while the drive down the freeway. If the computers get confused they call for help and stop if none is forthcoming.

    Missile firing drones have been flown remotely for years. A car/truck is harder because reaction times need to be faster. But a computer in the loop can solve that.

    That is the future. Outsourcing to cheap labor.

    • One operator in Mexico could monitor several trucks at the same time while the drive down the freeway.

      Meanwhile in Ontario, Canada, a girl gets a ticket for looking at her Apple watch while driving.

  • Securing America's Future Energy (Score:3)

    by 110010001000 ( 697113 ) on Thursday June 14, 2018 @08:19PM (#56787254) Homepage Journal
    Securing America's Future Energy is a lobbying organization that represents oil companies like Chevron. So this is just junk. Slashdot needs to stop reposting this drivel.

  • "We have a labor market characterized by churning -- continual job creation and destruction...The challenge is to make the transition as smooth as possible."

    Those displaced by new technologies on average do not recover back to the level they were. They take an economic hit. Similar applies to offshored careers.

    Therefore, just because new jobs are created by new technologies or offshoring, that does not mean people don't suffer.

    You are whacking one group to benefit another. Imaging robbing $1000 from 100 p

  • I'm a coder - and this is the first time I've heard that self-driving cars would be moving into my field!

  • Maybe in 2040, there will be no jobs, but in the 2020's driving as a job will be in a steady decline.

    This is just like printing over that last several decades. It hasn't gone away, but is a fraction of what it used to be.
    • I am at a division of a mega-corporation we have been a Cisco Gold Partner for years.

      Recently, we shut down most of our data center business because there's no value in it anymore. We sold off or laid off most of the data center workers and we were one of the biggest in the country a few years ago. What's funny is that 4 years ago, people said cloud would never take over... the developers apparently didn't agree and while no one in their right minds would us IaaS in the cloud because it's expensive and stup
  • Cars that can park themselves, and a kiosk for customers to select various services. Hertz has one of these two items already
  • how long
  • I look forward to the future where the car slams through the window of my office then sits at my desk to get some work done.

    -That, or when we turn into Mad Max world and have guitar flamethrower cars.

  • So these [dailymail.co.uk] are self-drining. And while I agree they won't take my job any time soon - after all, I'm a computer experts of sorts - I know for a fact that the ones shown already have taken ~300 jobs.

    As soon as it is economically 10x more feasible to do automated driving - be it little bots or human-transporting "car-like" things it will happen. Even if that requires standardizing street signs and perhaps some guidance system for automated cars. Transport is 70 million jobs globally. At least. The incentive to

  • If you drive a truck on long hauls over the desert, or in flat areas with decent areas, it's going to be sooner than you realize.

    If you drive a truck in regions with marginal roads, bad weather (snow, ice, avalanches, floods, etc), it will be a while.

    For examples take a look at the current lineup of trucks sold by PACCAR. More than half are now hybrids or fuel cell/electric, with some degree of self-driving automation.

Slashdot Top Deals

I have hardly ever known a mathematician who was capable of reasoning. -- Plato

Close