Home / General / The Jobless Future is Going to Be Great

The Jobless Future is Going to Be Great


If Democrats are going to start articulating pro-worker policies again as central platform planks, they need to get on board real fast to the problems that automation is already causing, problems that will grow rapidly.

Robot developers say they are close to a breakthrough—getting a machine to pick up a toy and put it in a box.

It is a simple task for a child, but for retailers it has been a big hurdle to automating one of the most labor-intensive aspects of e-commerce: grabbing items off shelves and packing them for shipping.

Several companies, including Saks Fifth Avenue owner Hudson’s Bay Co. HBC +0.82% and Chinese online-retail giant JD.com Inc., JD +1.79% have recently begun testing robotic “pickers” in their distribution centers. Some robotics companies say their machines can move gadgets, toys and consumer products 50% faster than human workers.

Retailers and logistics companies are counting on the new advances to help them keep pace with explosive growth in online sales and pressure to ship faster. U.S. e-commerce revenues hit $390 billion last year, nearly twice as much as in 2011, according to the U.S. Census Bureau. Sales are rising even faster in China, India and other developing countries.

That is propelling a global hiring spree to find people to process those orders. U.S. warehouses added 262,000 jobs over the past five years, with nearly 950,000 people working in the sector, according to the Labor Department. Labor shortages are becoming more common, particularly during the holiday rush, and wages are climbing.

Throwing nearly a million people out of work sounds pretty great! Hard to see any down side. Democrats should just offer some tax credits for employers to train workers. That will pretty much solve the problem!

Seriously, the real answer for this is going to need to be the right to a job, guaranteed by the government as an employer of last resort. I have no real problem with part of the solution being universal basic income, but again, I am extremely skeptical of Americans approving a welfare program that is not based upon work, as it files in the face of everything about American culture and history. Just the federally guaranteed job isn’t enough–free college tuition and the forgiveness of debt, a real industrial policy, the building of a green economy, and federal subsidies of everything from working in a farmers’ market to your local hipster bicycle shop are going to have to be pieces of the puzzle. People need work of some kind, even if self-defined. And they need a decent income and path to dignity. In a fully automated economy, they aren’t going to get it. And before someone says, “Derp, Luddite, Derp,” let me remind you that previous generations’ technological advancements worked because the increased jobs they created in an expanding economy absorbed those job losses. Today, we don’t create jobs to replace those lost through automation. They are just totally lost jobs, except for robot designers. Not solving this problem means massive social upheaval, no doubt channeled through racial violence, xenophobia, misogyny, and religious nationalism.

I have almost no faith that we will solve any of these problems.

  • Facebook
  • Twitter
  • Google+
  • Linkedin
  • Pinterest
  • DN Nation

    Baby Boomers peacing out to demand my generation just solve this is bad, though I know what’s worse: Baby Boomers themselves attempting a solution.

  • LurkinLongmont

    As a long time (25+ years) Automation Engineer, I’ve been wondering about this since almost day one. My bosses and clients have never given it a moment of thought. Maybe when we automate the guillotines they will….

  • GoBlueInSF

    The Senate Minority leader apparently has an answer. A tax credit for your boss.

  • Murc

    People don’t really think about this, either.

    I work as an IT support guy, desktop support. If you work in an office and have a guy on-staff whose job is to come by when your computer isn’t working and make it work again, that’s me.

    One of the big initiatives my current employer is trying to deploy within its IT infrastructure is a full-featured software catalogue, where you can one-click install any piece of software you might need, and if your computer isn’t working you can one-click hit a button and it will backup all your data, re-image the entire thing with a fresh version of windows, re-install all your stuff, and put all your data back.

    A lot of the other techs are really excited about this.

    “It’ll make our jobs so much easier!”

    “Guys. Bros. My dudes. You too, Erica. You know that one of the primary drivers behind this is that if and when they get it up and running, they can fire half of us, right? They’re not doing this to make our lives easier, they’re doing it because it’ll save them moneys.”

    “They wouldn’t… do… oh.”

    “Oh INDEED, my friends.”

    People don’t know until you point it out.

    • carolannie

      I keep coming back to…uh…hacking and all this IoT and totally automated production coming to a screeching halt because some bad actor decided to do something malicious. Even if you can automate your white hat hacking and security, there will be someone out there doing bad things,…and they will automate too. It will be a race to the biggest BSOD we ever did see. It would be ironic if our civilization crashed because of our AI. We certainly aren’t doing very effective work in isolating systems and protecting our automation.

      • carolannie

        A total systems’ crash certainly would create a lot of jobs.

        • wjts

          In the much the same way that the Black Death created many exciting new economic opportunities for the survivors, I suspect.

          • citizensuds

            Actually, it totally did. Peasant/laborer standard of living soared after plague, at least in late-Medieval Europe.

            • wjts

              Oh, I know. And I’m sure that was a great comfort to the millions who died.

              • citizensuds

                I wasn’t positing a trade-off. Your comment seemed to imply there wasn’t a strong improvement in subsequent living standards for the survivors, which, as a historical, empirically provable proposition, there was. It’s a bit counter-intuitive–you’d think losing that large a percentage of the healthy, working-age population would be an economic as well as a humanitarian catastrophe, but it wasn’t. At least, not in Europe.

                • wjts

                  No, I didn’t mean to imply that, though I can see how wedging my tongue into my cheek might suggest that I did. (Though I suspect – but could certainly be wrong – that in the immediate aftermath the medieval economy took a dive before bouncing back after things had settled down.)

                • Yestobesure

                  Wasn’t a big part of the improvement due to how shitty feudalism was? And that the deaths of so many serfs eroded the exploitative power of the lords?
                  Plus, an agrarian society is based on cultivating a fixed quantum of land, not on specialized division of labor, so massive population decline can easily raise GDP/capita in a way we wouldn’t see in a modern economy.

                • Banji Lawal

                  The peasants dots see their wages increase till landowners got the crown to pass laws forcing peasnts to accept the old wages that were lower. They also made it illegal for peasants to vote with their feet and go to other places or the cities. Then when the serfs did oppose these interventions in the “free market” they were killed, imprisoned, and such. So in the end it kinda became a wash with the nobility, church, and landowners encroaching on the rights of people who worked the land.

                  And in IT when there’s a disaster it doesn’t mean they’ll have to hire the same amount of people. They’ll get by with a lot less

                • firefall

                  Well, it was more of a redistribution, peasants and townsmen got a lot better off, a lot of the nobility and gentry had a big drop in income. Poor things.

              • Matyas_L

                Picky Picky Picky. You gotta make lemonade out of lemons, right?

                • wjts

                  Can’t make an omelette without killing 30-60% of your chickens.

          • sibusisodan

            Learne the one weirde tricke for cutting your seigneural dues. Leige lords hate it!

          • carolannie

            *grim laughter off-stage*

    • SatanicPanic

      Sure, but if your employer is typical it’ll be 10,000 years before it works well enough to actually fire anyone.

      • Murc

        Especially with some of us actively sabotaging it!

        • addicted4444

          And then they will go all in on the cloud and BYOD and sorry, no more jobs, nor any sabotage opportunities.

          It’s a little frustrating hanging out at Atrios’s place these days because his schtick seem s to have become to simply presume that self driving cars won’t happen (except now that they are happening, the goalpost shifting has begun, from it being a technical impossibility, to it being too dangerous, and now we seem to have come to it will happen, but only because governments will allow it to happen by forcing changes on people).

          Automation and the jobles future is almost upon us. By pretending that it’s not gonna happen, liberals are, once again, losing the opportunity to prepare for it.

          I appreciate that Erik has actually started recognizing this, because he seems to be one of the few on the left that is giving it decent thought.

          • guthrie

            I kind of thought the whole point was that our jobs are automated nearly out of existence, then we rise up and seize the means of production.

            Of course, someone has pointed out that the rise of autonomous killer drones rather damages this possibility.

          • applecor

            My experience is that a lot of people are giving it decent thought, but are not coming up with any answers. This thread has pretty well laid out the counterargument to just about every conceivable policy.

  • Scott P.

    let me remind you that previous generations’ technological advancements
    worked because the increased jobs they created in an expanding economy
    absorbed those job losses. Today, we don’t create jobs to replace those
    lost through automation. They are just totally lost jobs, except for
    robot designers.

    I am pretty sure this is not true. Do you have any citations for that?

    I mean, it’s fair to say that automation depresses wages, that it leads to disruption, that the gains are unequally distributed, sure. But the jobs go poof? No.

    Moreover, it wasn’t just lucky coincidence that previous episodes of automation were linked with an ‘expanding economy’. Increasing productivity is what led to economic expansion.

    That’s not to say laissez-faire and everything will be fine, before someone tries to put words in my mouth. I’m on board with most of the proposed policies. But we need to diagnose the problem correctly if we’re to develop good solutions.

    • sibusisodan

      I guess the problem is that the pace of new job discovery probably won’t keep pace with accelerating automation. That and income distribution. Both big problems.

      To say that there won’t be new jobs discovered/invented entirely seems rather absurd.

    • I’ve wavered recently on this. (Fully confession: I’m an automation engineer).

      To even say automation has historically depressed wages isn’t really correct. If productivity is increased, you can lower prices, demand increases, and production should expand to meet the higher demand. Jobs may go up more slowly than they would with an identical increase in demand without the productivity boost, but its hard to see how jobs would be eliminated in raw net, let alone where such a demand boost would come from. The period when gross production grew but wages stagnated was in the 19th century when automation was primitive (and child labor was everywhere!). Productivity growth was low.

      In service industries, where demand is more constrained, there could be a problem. But I don’t see any way to economically automate service industries in a manner that would really cause problems; it requires to much behavioral dexterity to perform these things to be feasible to automate in a way which is cheaper than even having a decently paid human do it.

      I’m not saying that income inequality isn’t a huge issue, but I don’t see automation as making it much worse in the foreseeable future. To my mind it’s a political problem, not a technological one.

      • Tim Reynolds

        I’d suggest not reading too much into a ‘history’ that essentially consists of only 150 years of significant technological advancement within single lifetime, at least when it comes to mechanization.

        A whole lot of people want to treat a century or two as if that’s a much, much longer period of time than it actually is, and is if it’s a lot more representative of the over-all human experience than can be supported by the anthropological record.

    • Yestobesure

      We need to be serious about running the economy hot so unemployment stays low- no premature tightening by the Fed, real catchup spending on our infrastructure deficit, and an eye to supporting state and local budgets so they can deliver labor intensive services like education.
      Erik’s idea of subsidies for work sounds like beefing up the EITC, an excellent idea.
      If the economy is running well, people will be able to find jobs. For some people, they may be lower paid jobs than what they lose to automation, which is why supplementing income and beefing up the safety net is a good idea.

      • sanjait

        *stands up and slow claps*

        I will add, not only does a hot economy keep employment low, but keeping employment low has the huge secondary benefit of forcing employers to hire and train more marginally qualified workers, and to train and promote its existing workforce more aggressively.

        In other words, it incentivizes companies to help workers acquire in demand skills.

        • billcinsd

          I will add, not only does a hot economy keep employment low, but keeping
          employment low has the huge secondary benefit of forcing employers to
          hire and train more marginally qualified workers, and to train and
          promote its existing workforce more aggressively.

          I think you want to keep unemployment low, not employment

    • Gepap

      “Increasing productivity is what led to economic expansion.”

      This statement is both correct and misses the point entirely.

      “Economic growth” is a measure of increased production. Obviously, if each unit of production can produce more with the same given inputs (ie. productivity) then that leads to growth. The fallacy lies in conflating the unit of production to a human laborer.

      The shift between an agricultural economy and a manufacturing one wasn’t “just coincidence” but I think this misses the real forces at work. Increasing productivity in the farm sector led people out of the farms. This created a lot of jobless people, which gave industrialists the opportunity to grow, since they now had a displaced and desperate workforce which came cheap. Since at the time the only option was to use people, well, there you go. Manufacturing never really came even close to replacing all the jobs lost in the Agricultural sector (a sector that at one time employed 90% plus of workers. Manufacturing never came even remotely close to having 90% of the workforce). General economic growth (led by the energy revolution more than anything) increased wealth, which allowed for a vast expansion of the services sector, which was the sector to absorb most of the labor force. Again, given the level of technology, human labor was the only option, so, hurrah….

      Now this has changed – human labor is NOT the only option for an increasing number of tasks. Human Labor can be replaced by additional Capital (in the form of productive machines). There are only really three sectors (Agriculture, manufacturing, services). The need for humans in the agricultural sector has been greatly minimized, and even now remains in decline as jobs still left mainly to humans (vegetable and fruit picking) are starting to be mechanized. The need for humans in the Manufacturing sector continues to decline, leaving only services. Once services goes, there is no sector left.

      The only avenue for growth of paid labor that I see is an expansion of the Service sector, by turning activities that are currently done outside the labor market and monetizing them. This means replacing at home care by relatives with paid caretakers, for example. But even this has its limits. At some point in the technological sprectrum, a human is simply not the more economical unit of production anymore, for almost anything.

    • so-in-so

      Yeah, there were numbers of older farmers who lingered on the land and never made the transition to the industrial economy. There are stories in the North East of the old hill-town farming communities collapsing as the younger and more adaptable people moved to the industrial valleys to escape. In our modern case, I’m sure not everyone is suited to getting a degree and taking a tech job, even if there were enough such jobs, even if you paid for as many years as needed to graduate.

      • BigHank53

        There are plenty of ghost towns and abandoned homesteads up there. If you hike the Wapack trail you will pass an enormous stone wall made of granite, four feet tall and three feet wide. The farm it surrounded has been gone since the Civil War.

  • SamR

    I can’t say I’ve figured out an exact solution, but it does seem off that we have one group of people working extra-long hours and others who don’t have any work. I’ve experienced both of these circumstances, both are uniquely terrible in their own way.

    • JamesWimberley

      Efficiency wages. The theory also explains long hours.

  • sanjait

    This isn’t truly new. Warehouse automation has been a big thing for some time.

    Ironically, warehouse employment has nevertheless increased because shopping went online, pushing demand up faster than automation could offset. This is coincident with the rapid collapse of retail.

    And warehouse automation is just one example of automation replacing workers across the economy. This is, far more than trade, the major driver of decline in manufacturing employment in the United States.

    The remaining question is what to do.

    There are two high level strategies often discussed. One is to help workers attain the skills that are in demand in the modern economy, through support for primary education, higher ed, lifetime learning and other mechanisms of skill development. The other is to make the welfare state more robust.

    The situation is urgent enough that both of these should be heartily supported.

    A third strategy, rarely discussed, would be to enact policies to promote migration within the United States, as the geographic distribution of labor demand is very uneven.

    Actually there are two other strategies often discussed: trade protectionism and immigration restrictions. But I forget these because they are red herrings that have no hope of meaningfully addressing the issue at hand.

    • NeonTrotsky

      “One is to help workers attain the skills that are in demand in the modern economy, through support for primary education, higher ed, lifetime learning and other mechanisms of skill development.”

      Massive numbers of retail workers and truck drivers just aren’t going to become software engineers, partially because they might never have the skills/don’t want to be software engineers, and because even in a fully automated future there still won’t be enough of those jobs for everyone. What do we do about this population?

      • Retail workers are getting displaced because of Amazon, not automation, right? (seriously, I don’t know).

        Truck drivers are going to be fine, except for their wages dropping for reasons not related to automation: mostly lack of unionization and corporate greed. its a much longer way off than the venture capital leeches in silicon valley mags try to get people to believe. If it were reasonably technologically and economically feasible, why do trains still have engineers and conductors?

        • NeonTrotsky

          I mean Amazon is pushing automation unless I’m mistaken, their entire interface is essentially a replacement for visiting a physical store, and I believe huge parts of their warehouses already are at least partially automated. Software displacing human work is as big of a problem as hardware displacing it, for example some experts think Paralegals will be put out of work within a decade or two, purely because of software advances. I think a lot of white collar workers are going to be in for a shock when they realize they aren’t immune to this trend either.

          • sanjait

            This. “Automation” really should be thought of broadly, including any kind of machine function replacing human functions. This is the world we live in, it is changing quickly because of robots AND computers, and the pace of change is probably accelerating.

          • I don’t know. I had a buddy who worked in an Amazon warehouse a few years ago (stopped in 2015) and it sounded like it was pretty labor intensive. Maybe they’ve had more sucess in automating recently, but they’re still advertising heavily for jobs at the local warehouse. Bots capable of picking would either require a complete re-engineering of the process, or crazy expensive (and very breakable) C3PO’s. It’s hard to see how any of those options would be cheaper than paying guys $15-18 an hour.

            A lot of what Paralegals and the like do is data curation, otherwise it’s hard to see how google hasn’t replaced them already. Curation is something that’s almost impossible to automate, that’s why netflix and youtube are so crappy at suggesting music and videos, and even they are piggybacking off of human curation.

            I feel uncomfortable being the optimist in any discussion, but again, I think our problems are more solvable than techno-jobless-apocalypticism would imply.

            • applecor

              Paralegals should be drafting a lot of documents that lawyers currently draft, just by making menu choices. The technology has been around for 20+ years, but since law firms bill by the hour, it hasn’t caught on. It will.

              • If its been around for 20 years and hasn’t been adopted, what’s going on? What makes it so certain its going to catch on now?

                If a simply policy like operating under billable hours can prevent automation eliminating a profession, that doesn’t make it look like we’re going to hurdle towards joblessness anytime soon.

                • applecor

                  It will catch on eventually because law firms do face competition on price. I view it as a temporarily inefficient market. It hasn’t caught on so far because it basically requires law firms to make an investment they have not had to make previously, so they can cut their prices – not a good look, on the surface.

                  There are cultural barriers also. People that do routine documents like wills and residential real estate transfers already use this software, but BigLaw continues to believe that its special-snowflake 100-page contracts need an associate slaving over commas at $400 an hour. Eventually the clients will make them get over it. When they do, the market for junior BigLaw associates will contract somewhat.

                • (((realinterrobang)))

                  Wills are not routine, even though they’re common. There are so many variables involved with creating wills that any menu-driven boilerplate generator for wills would have likely hundreds of choices. A lawyer friend hired me (in my capacity as a technical writer) to see if we could design a “Will Creation Kit” that was basically a set of worksheets his clients could fill out so that he had all the information to do a will in one shot, and it was way more complex than you’d figure.

                • applecor

                  I’m pretty familiar with the technology involved. Sure, a will has hundreds of choices, and the initial programming, even for a short document, is huge effort, but exactly BECAUSE wills are common, it is worthwhile for someone to program it, because you’re getting more leverage out of the effort. Some things might only be leveragable over a single firm, but the more common documents could be generated via off-the-shelf software for sale. Sure, you need to handle 50 states etc. but TurboTax seems to handle that OK.

                  I actually saw a 200-page debenture produced with software like this. That was in the early 90s.

              • Matyas_L

                I teach in department with a Paralegal/ Legal Studies program. We train a lot fewer Paralegals than we did twenty years ago. The major slumped in the 2007-08 recession and has not really recovered. It used to be able to weather those economic downturns and it doesn’t anymore.

              • Drew

                A really good paralegal can do everything a lawyer can do except for the stuff you legally require a law license for, like taking a deposition, appearing in court, signing a pleading or motion (although they can draft it).

                • applecor

                  Yes – the problem is that “really good paralegals” then go to law school, ruining everything.

                • Drew

                  Well, not all of them. You seem to have basically two kinds-the fresh out of college kids looking to go back and career paralegals. There are definitely a fair amount of really good paralegals who aren’t 20somethings looking to bail/go to law school asap.

                • applecor

                  In 25 years at two BigLaw firms, I never saw (corporate) paralegals used properly or developed. The few career paralegals that the firms deigned to keep around were certainly very helpful, but if the firms could bill out an attorney for something a paralegal could do, they generally would. It seemed like the only time paralegals were trusted with anything substantive was when some completely burned-out associate simply had no other choice, either because of a time constraint or because s/he simply didn’t know how to do the ask, and the paralegal did.

            • sanjait

              At present, machine learning occurs at the level of fairly simple tasks. Trainers throw gobs of examples at a system and it learns to come up with correct solutions.

              But they don’t synthesize even moderately complex solutions well. So they can’t replace much of what humans do.

              At present…

              • “But they don’t synthesize even moderately complex solutions well. ”

                Tell me about it. But, speaking from my own experience, the jump from simple, functional learning like we have to complex adaptive learning like we’d need is bigger than between horse drawn carriages and Doc Brown’s Plutonium charged flying Delorian.

                • sanjait

                  It does seem large. The pace at which machine learning erodes that gap is, to me, a huge variable in all this discussion.

                  Like you I think the gap is wide and it will take quite some time before machines can really think like humans.

                  But … I could not confidently estimate. 10 years, 20 years, 50 years …? From what bit I know about the technology, it both surprises me with advances in some areas and with its primitiveness in others.

            • SatanicPanic

              Amazon is unusual though- they have warehouses full of random stuff going out on a random schedule to random places. A warehouse full of shoes or TVs will have less trouble cutting workers.

            • so-in-so

              So basically you dispute the story that the OP is based on.

          • for example some experts think Paralegals will be put out of work within a decade or two

            OMG, what will JDs do then???

        • Eric K

          “Truck drivers are going to be fine”

          the first people displaced will be truck drivers. The potential labor cost savings for trains isn’t as great, once the technology is there I’m sure they go for it, but they don’t seem to have the need to invest in it and be on the cutting edge.

          I think this is one of those technologies that seems way far off but then once it starts will become ubiquitous way faster than anyone ever imagined.

          • Most people tend to think it’ll be exponential like you say. My own best guess is that the learning curve is logarithmic: we’ve captured the big boost already and now its going at a crawl.

            We’ll see though.

            *Side note– if the labor cost is so much smaller for trains that we haven’t automated them (a much, much easier task than automating cars and trucks), why don’t we use them way more, unless labor cost is not a big factor in shipping. In which case why is automating trucks considered such a lucrative and worthwhile endeavor? If any economists could fill me in I’d appreciate it.

            • sanjait

              Just an educated guess: the portion of the per mile shipping cost that is accrued by the driver/conductor is much higher for trucks than for trains.

            • Gepap

              Trains do in fact ferry vast amounts of materials. They haven’t been automated because that automation would require mainly track and signal investments, which would be immensely costly, and shipping is not a very high margin industry so rail companies don’t have huge sums lying about to do this. On the other hand, it is the manufacturers of trucks themselves that have an incentive to create automated vehicles for their consumer base, which gives them an incentive to innovate that does not fall unto the actual shipping companies.

              Also, trains do ship vast amounts, but they aren’t flexible. You use a train if you want to ship vast amounts between two points connected by tracks. If you are shipping anything to a place not next to tracks – well, you need a truck.

              • “On the other hand, it is the manufacturers of trucks themselves that have an incentive to create automated vehicles for their consumer base, which gives them an incentive to innovate that does not fall unto the actual shipping companies.”

                If their consumer base is shipping companies, what is the incentive to innovate by creating a higher cost product that doesn’t lower their customer’s costs by much?

                • Gepap

                  For a trucking companies, the drivers are a huge expense. You have not pay them wages and benefits. In addition, you have limits imposed on the hours they can work due to safety reasons. What gives you the idea that labor is not a big issue for trucking companies?

                  Without drivers, shippers could increase the number of trucks on the road, have them moving nearly 24-7, and never worry about sick hours or disability/workers comp.

                • If they are such a huge expense, why is long range trucking a thing? Why don’t we ship all consumer goods via track and short range trucking from depots?

                  I don’t know, something in the equations just doesn’t balance here. Maybe the massive subsidies the interstate highway system provides account for it (I’ve heard guys at my dad’s train shows argue that), but again that makes policy a much bigger driver than technology. And as above, that leaves some healthy room for skepticism on the policy level even if you’re an optimist about the technology

                • Gepap

                  Where to start:

                  1. The freight rail network has plenty of holes – so lots of locations are not covered by the network which would not allow for some endless hub-and spoke system. The road system has much greater coverage.
                  2. Rail stock is not infinite – the system has a finite carrying capacity.
                  3. That carrying capacity means that pricing wise, the system makes most sense if you can ship many container’s worth of cargo between two points. If you need to ship a smaller quantity, even over distance, it can be much less of a hassle to do so individually.

            • Eric K

              We do use trains a lot. They are really good for getting a lot of stuff from one point to another where there are tracks. Where trucks come in is once the train gets to its end point the trucks disperse everything to the final destinations or trucks are for smaller quantities.

              • And those final destination trucks are going to be the last ones to automate. In fact most proposals for any truck automation I’ve seen all involve long range trucking being automated, with short ranges still being done by live drivers.

                I think under-girding this is that we assume that the infrastructure investment which would be required will be born by the government for automated trucking (along with some very sugary optimism about just how easy the process is to automate to begin with). But a similar thing hasn’t happened with trains, which would be much cheaper. Its hard to see why it would happen for trucking unless we just get some Apollo program style enthusiasm for the project. And that’s assuming that the technology keeps up with the cheery projections that we keep assuming.

                • so-in-so

                  One train can haul the equivalent of several hundred trucks(so the savings by automating are much less), but only between defined places where the tracks (paid for by the RR company and taxed as property) run. The cost in getting this freight to off-line customers comes in moving from the train to the truck. This has been minimized to some extent with trailer-on-flatcar and containerization, but is still an expensive and time consuming process. The RR infrastructure has contracted (did I mention the RRs own, maintain and are taxed on it?) so the distances and costs increase for many shippers and receivers. If you can fit it in a truck, and have to drive to another city anyway, shippers use trucks.

                • But that means that operators aren’t the driving cost, infrastructure is. Any feasible shift to automated trucking would require massive infrastructure investment (or Skynet level AI breakthroughs).

                  If the cost structure is such that train and long range truck shipping coexist, then its hard to account for why we could presume a massive increase in infrastructure investment is worth the payoff in reduced driver costs for trucking, when simply shifting to trains would lower driver costs without any corresponding investment on the shippers part. Maybe the actual numbers do bear this out, but I have a feeling that there’s some unreasonable techno-optimism or the assumption that someone else is footing the infrastructure investment bill going on here.

                • liberalrob

                  I think the issue is that it’s much easier to put a few more trucks on the publicly-funded highways than it is to lay down more privately-funded rail lines. If the highways get beat up and need fixing or expansion, some or all of the cost is borne by the public sector. Also, adding more trucks to a highway can be done quickly and incrementally; putting down a rail line that isn’t used to capacity is wasted potential.

                  Given that, I think there will be a lot of pressure to create “automated truck only” lanes on or adjacent to existing superhighways. Of course they would be the functional equivalent of rail lines, but they could be sold as getting a lot of those 18-wheelers out of the non-commercial traffic lanes while utilizing existing right-of-way.

                • BigHank53

                  When’s the last time you saw fresh rail being laid? There’s 140k of rail in the US. In 1890 there was 129k. Capacity has gone up and trains have gotten bigger, but many rail corridors are operating at 90+ percent of maximum capacity.

                  Also, don’t forget the very high costs of loading and unloading cargo: 40′ shipping containers are magic. Moving pallets of something from a boxcar to a truck would probably cost more than you’d ever save in freight costs. (Though robot forklifts may tweak the math on that.)

                • Provided the government is picking up the tab on the infrastructure, trucks are plainly better for most shipping. The issue here is whether the government will pick up the tab for automation– because you’re not going to see self-driving trucks en mass on our current highway system.

                  Still, automating rail is a senior at MITs control engineering project, you have a track, switch, throttle, path. Its hard to see why it hasn’t happened unless the economics don’t line up. If the economics of rail are so different from trucking, its hard to see how the market has stabilized in shipping the way it has. If you add in the extra cost of infrastructure upgrade (which would be much greater for trucking than rail–which really wouldn’t only require fitting the trains with some cameras and networking up the switches–which they’ve probably already done) and the necessary technology advances, I don’t see truckers being replaced within anywhere near the time horizon some are predicting.

                • so-in-so

                  The major railroads which would be easiest to automate – unit trains of coal from mine to power plant, unit container trains from port to distribution point – are also still strongly unionized. The non-union railroads are the smaller regionals which, like the last-mile truckers, are harder to automate.
                  RRs are also VERY conservative in the older sense of the word. One CEO described the culture as “any change someone suggests is met with ‘they tried something like that in 1906 and it didn’t work, so don’t bother'”.

      • sanjait

        Software engineers isn’t the only in demand job in the modern world. The jobs categories with growing demand are broad, and including many technical positions like software devs and data analyst, but also positions across the healthcare domain, and even a range of blue collar trade skills like plumbing and electrical.

        So to start, we should remember that the economy is complex and dynamic.

        We should also remember that the problem is a big long term one. It’s critical that we prepare the current generation of children for the modern world, so they don’t get stuck in the same low skill trap as many of their parents.

        And while many of those parents lack “train-ability” for many modern jobs, a great many of them are train-able in something useful.

        If we are able to shift the skill level of the workforce broadly forward, that in turn increases demand for lower skilled workers. Rising tide, in the 60s sense…

        • Murc

          It’s critical that we prepare the current generation of children for the
          modern world, so they don’t get stuck in the same low skill trap as
          many of their parents.

          Many of those parents absolutely have skills. They have skills they’ve developed over many years, in fact. They got jobs with those skills. They fulfilled their half of the social contract.

          It’s society that failed them by telling them in their forties “sorry, we don’t need your skill anymore.”

          And while many of those parents lack “train-ability” for many modern
          jobs, a great many of them are train-able in something useful.

          That’s lovely. Will we be paying their mortgages and giving them a living stipend while they re-train? Because that, in my opinion, is the bare minimum they’re owed. If we automate someones entire profession out of existence, they are OWED.

          • sanjait

            Yes, but when did merely complaining that the world is unfair and unlovely solve problems?

            Notice how I explicitly endorsed enhanced welfare protection as one of the two main strategies for addressing the issue at hand.

            So what exactly are you arguing?

            • Most people don’t want welfare or to be wards of the state. They want to feel like their lives have some meaning, and that that meaning can translate into a stable job and career that allows them to settle down, buy a house, start a family, and so on. This is one of the biggest complaints that people have about getting subsidies. They don’t want subsidies. They want jobs that pay well enough for them to afford stuff on their own.

              And who is going to pay for this welfare, anyway? The rich who are busy replacing these workers with machines? The rich who essentially control the levers of government to the point that they are on the cusp of being able to toss millions of people off health insurance just so they can get a tax cut? Do you think they are going to agree to any of this?

              • sanjait

                I think it’s useful to pencil out possible strategies for addressing problems. Tradeoffs and political challenges are inherent and should be acknowledged, but at this point I just feel subject to a Gish Gallop and I have no idea why.

                • The robot apocalypse is a touchy subject in these parts ;-)

                  That being said, I’m all for discussing solutions. But realize that we have a government mostly controlled by a political party that is committed to eradicating people who need welfare, not in increasing the number of people on it.

        • And if my mama had wings she could fly.

          Once the owners of capital own semi-intelligent machines they will own labor. Once they own labor, they won’t need people to make their luxury homes and yachts. They won’t need people to make their designer clothes. They won’t need people to dig up and process the raw materials. They won’t need people to treat them when they’re sick. They won’t need people to drive and fly them around. They won’t need people to fix their machines. They won’t even need people to protect them. They won’t need people, period.

          What do you think will happen when some enterprising Trump jr. sociopath decides that the world could really do with a whole lot less people milling about griping about not having anything to do?

          • sanjait

            I think its justified to worry about that future, but not at the expense of thinking how to address the issue of workforce dislocation in the current generation. The former is an existential threat to the human condition but the latter is still a big deal and urgent issue. No?

            • Workforce dislocation is not the problem. Workforce disillusionment is the problem. People are not just widgets that will happily be retrained for whatever shit jobs will remain in the new economy. If that’s the immediate future, then we might as well start ramping up opiate production, because there’s going to be a whole lot more hopeless people out there who will have nothing to gain or lose in their lives.

              The sad thing is that this is not in any way inevitable. It is merely taken as inevitable because the people developing the machines say it is inevitable.

              • sanjait

                “Workforce dislocation is not the problem. Workforce disillusionment is the problem.”

                What caused the disillusionment? Dislocation. This seems plain.

                “The sad thing is that this is not in any way inevitable. It is merely taken as inevitable because the people developing the machines say it is inevitable.”

                It is inevitable. Stopping the development of smart, efficient machines is not possible under any policy regime. I can’t imagine what you think could change that.

                • It is inevitable. Stopping the development of smart, efficient machines isn’t not possible. I can’t imagine what you think could change that.

                  You want to know what will change that? A whole lot of pissed off people armed with clubs and Molotov cocktails. Just wait. It will happen.

                • sanjait

                  Yes, and then what? Does the Luddite strategy (and yes, that word just became appropriate) do any good?

                  I think you are outraged, and arguably rightfully so. But I didn’t cause this problem, and it seems like we are talking past each other.

                • I think if we focused research on developing technology to assist people instead of replacing them, then a lot of these problems would be solved. We don’t need driverless cars. We need cars that help people drive better. We don’t need factories filled with machines. We need factories filled with people assisted by machines. And so on. But the root problem is not the technology; its the commitment to efficiency at all costs and the belief that machines will be better than people. That is what needs to be changed.

                • But I didn’t cause this problem

                  You just need some job retraining; once you learn to cause problems, the world will be your oyster!

                • sanjait

                  I could work for the Russian government in the “active measures” department…

              • Yestobesure

                People need something to do to feel useful. That something doesn’t have to be work in the sense that it exists today. Lets face it, many people are miserable in their jobs. If they could spend more time caring for family, being creative, learning — and still maintain living standards — that’s a huge win. If technology makes it easier to fulfill human needs with less work, that’s possible.
                Maybe we have govt pay people for riding stationary bikes (Black Mirror reference) or better, for commenting on blogs.
                This is way out in the future, and besides the economic and political we will be dealing with massive social change… but interesting to imagine.

        • Gepap

          The first issue with this is that if “everyone rises a skill level”, wages for skilled jobs collapse. The reason electricians and plumbers get paid well is because they are rare. Make trained electricians a dime a dozen and it becomes a minimum wage job.

          What upper “skill level” will come next? Make everyone surgeons? Some jobs are rare cause demand is low as well as it needing high levels of training. Missing from this “there will always be jobs!” fantasy is that “a job” is not just activity. Its paid activity – it is someone able to rent their times and skills out. The question is what are they competing against. Right now for a lot of things, humans remain the only game in town. Great – the problem lies in thinking this will always be the case.

          Moore’s Law has remained relevant for some time – issue with Moore’s Law is that it is an example of exponential growth. Humans are terrible at envisioning what this means. Since the inception of the computing revolution, creating computing power equal to our brain power has been out of reach because of how great the disparity was. Except that, as computing continue to grow exponentially, the moment at which that immense gap becomes not so immense and then disappears shrinks from a couple of generations to a couple of years.

          • sanjait

            “The first issue with this is that if “everyone rises a skill level”, wages for skilled jobs collapse. ”

            That’s not how it works. If it were true, the post-war period in the United States, when educational attainment shot up dramatically, would not have been broadly prosperous.

            • Gepap

              “That’s not how it works”

              The Law of supply and demand has not been revoked, so yes, this remains how it works.There is no way you could claim that a tripling of the number of electricians out there would allow wages for electricians to remain unchanged.

              According to the census 42.5% of the population has an associates degree or higher, only 32.5% has a bachelors or higher, and 12% have a professional degree of higher. So even today, a minority of the over 25 population has a degree of any kind. And having a Bachelors doesn’t actually mean you have been trained at something. A BA in Political Science for example does not specifically qualify you for any specific kind of job. It merely denotes you have achieved a certain level of academic training. What jobs such a person ends up at varies. A lot of BA’s are like that – educational signals but not something tied to a specific job.

              You are conflating the growing prosperity in the post war era with an increase in education, but this is not really accurate. Growth in prosperity n the US has been driven by basic increases of overall productivity, some of which are related to higher education, some of which are not. There are been in increase in demand for college graduates, which is what is driving people to get a degree, but again, a lot of these degrees aren’t specific to any job.

              • Yestobesure

                Which is why everyone is better off in low skill countries than high skill ones.
                Yes in individual fields, an influx of skilled workers will lower the wage premium. That doesn’t mean that everyone is worse off when labor acquires more skills. Lower wages for electricians lower the cost of living for everyone else. Think macro, not micro.
                We just need to keep the macroeconomy humming, make the safety net stronger and fight entrenched monopoly power.

                • Gepap

                  “Lower wages for electricians lower the cost of living for everyone else. Think macro, not micro.”

                  Presuming that the same dynamic isn’t at work in whatever job sector the non-electricians work at. You are failing to think macro if you aren’t realizing that greater automation will create certain work ghettos for human beings.

                • Yestobesure

                  More people doing non-electrician skilled work will lower the cost of living for the electricians. We all serve each other, and the better we collectively get at serving each other the better off everyone is. Presuming that all the gains don’t accrue to holders of capital/monopoly/IP.
                  What do you mean by “work ghettos”? Not familiar with the term.

                • Gepap

                  ” We all serve each other,”

                  No, we don’t. We all participate in a single economy. Those two things are not even remotely synonymous. The question is what kind of commercial exchange you can enter into with others. Your last presumption is the key – what allows most of us to get any share of the surplus is that for the economy to work, human labor remains a necessity, and thus individuals can commercially rent their time out to get their share. At a certain level of technology this is no longer the case. The holders of the capital can replace labor with more capital. Having to pay a human to produce a good or service becomes unnecessary.

                  The questions at hand are how long until we get there, and what happens socially once this one avenue for people to get a share of the overall product disappears.

                • Yestobesure

                  Labor share of gdp is 60% (would that it were higher!). That means a very large portion of your spending is other people’s income and vice versa. It also means we’re a long way off from replacing all humans with machines.
                  It’s hard to picture the future, but a key question will be how tightly held labor saving technology is. If ownership is diffuse, competition should move the benefits of higher productivity to the consumer, not to capital or the IP holder. Real wages rise (on average, and again maintaining a low slack macroeconomy to prevent unemployment). If instead we have oligopoly control of the new means of production… that’s bad.

                • Gepap

                  “It also means we’re a long way off from replacing all humans with machines.”

                  Massive change can happen in historically brief periods. Just look at the changes between 1850 and 1950 in living conditions and situations for residents of this country. Any technology sufficiently game changing will upend systems fast.

                  You also don’t have to replace all humans with machines to cause problems – if the labor share of GDP went to say 30%, you still have massive societal problems.

                  And yes, how widely ownership of the means of productions are shared will matter greatly to the outcome.

                • sanjait

                  You’re better at explaining this than I am. Cheers.

              • sanjait

                “The Law of supply and demand has not been revoked, so yes, this remains how it works.”

                No, it really isn’t. The economy isn’t one with just electricians and a fixed demand for electrician work. Not even close. Supply and demand remains in effect but the curves move in a dynamic system. Demand for neither total labor nor specific job titles are not fixed. Look up “Lump of Labor fallacy” if you really want to get economics-y.

                “You are conflating the growing prosperity in the post war era with an increase in education, but this is not really accurate. Growth in prosperity n the US has been driven by basic increases of overall productivity, some of which are related to higher education, some of which are not. ”

                Yes, I am conflating the two, because they are linked, as you acknowledge. You are right, it’s not quite that simple, but neither is the coincident rise in productivity and educational attainment in the post-war period merely a coincidence.

                One key measure I think is important is the college wage premium. We often hear complaints about how college degrees don’t prepare people for the modern working world and don’t help people get ahead, but actually, the college wage premium has been rising, not falling. It is now *more* valuable to get a college degree than ever. It doesn’t matter that degrees aren’t typically tied to specific jobs, they help people get jobs and employers think they make them more productive. Are those employers all wrong in your opinion?

                • Gepap

                  I would argue that the “college wage premium” and the fact its increasing speaks directly to the weakness of labor’s overall bargaining power.

                  Employers have the power to demand that large cadres of workers get a college education before even considering hiring someone. Being able to get a degree denotes certain ‘desirable’ qualities – intelligent enough to take semi-complex instructions, a certain level of intellectual flexibility, a certain level of intellectual discipline. This is what makes a “degree” valuable in an of itself. If the “premium” is increasing, it is because even now large percentages of the population aren’t willing or able to meet these demands, which immensely depresses their bargaining ability.

                  None of this is good. At a minimum, half the population is shit out of luck. And supply and demand does remain – if you saw a vast expansion of the people who were able to get bachelors (say, up to 60% instead of 33%) then overall wages for these folks would decline, and the talk would be about the “increasing premium for Master’s degrees”.

                • Yestobesure

                  The classical economic argument is that as a higher percentage of the population gains skills, supply and demand lowers the skill premium and makes the remaining unskilled workers more employable and better off. I will admit that we haven’t seen this in any real way.
                  To me this is the best argument for min wage, labor laws, and unions – increasing bargaining power for the low skilled at a time when they really need it.

                • Gepap

                  ” remaining unskilled workers more employable and better off.” No, that is not what i said at all. I believe I specifically stated that this sector of the population is “shit out of luck”.

                  And yes, we do need to raise labor’s overall bargaining power, so stronger unions and a decent minimum wage are necessary. I would argue that all workers, not just low skilled workers, benefit from being in a union.

                • Yestobesure

                  Wasn’t trying to restate what you said – was giving an alternative view that hasn’t held up well to date.

                • sanjait

                  That’s it, but it’s a ceteris paribus assumption.

                  I truly believe that is how things work, but that the slowing increase in educational attainment and other big secular shifts in the nature of work have more than offset that effect.

                • sanjait

                  I don’t really get that argument.

                  I buy that labor has weakened bargaining power in the modern world (for reasons … which would create a whole new discussion if we started discussing).

                  But why should that affect workers differentially? Just because employers can demand workers get a college degree, why, in a world of increased employer bargaining power, would they not *also* shit on college educated workers too?

                  You’re totally right though about what a college degree means in the modern workplace. It denotes desirable qualities … that are important in the modern workplace. AKA, things that employers think make people more productive. And I think those employers are not in aggregate wrong.

                • Gepap

                  “But why should that affect workers differentially? Just because employers can demand workers get a college degree, why, in a world of increased employer bargaining power, would they not *also* shit on college educated workers too?”

                  Because College Educated workers remain a minority of the workforce, so you still have a “scarcity” issue.

                  Maybe, instead of thinking of the “college premium”, why not think of it as the “high school penalty”? How much less workers who don’t meet the “get a college degree” demand suffer for their failure.

            • so-in-so

              Well, yeah, the missing trick is getting the rest of the industrial world to destroy itself (without actually blowing up the entire thing with nukes).

              • sanjait

                No, that’s not why the US had a major increase in domestic productivity, educational attainment and broadly shared prosperity at the same time.

                I know that’s a commonly told story, that we were only doing well because the rest of the world was destroyed, but that’s not really what happened.

          • Humans are terrible at envisioning what this means.

            Hell, humans are terrible at envisioning polynomial growth. More precisely, they are terrible at envisioning polynomial decay; as a friend and collaborator pointed out at an AMS special session once, when mathematicians sketch even a fuckin’ parabola y=x^2, much less the graph of y=x^n for n>2, they entirely misrepresent how flat it is near 0: when x decreases to 0, y decreases to 0 much faster. Like, (1/10)^2 is ten percent of 1/10, but (1/100)^2 is one tenth of one percent of 1/100, and so on. (So, going the other way, the growth of y=x^2 as x increases from 0 is nowhere near exponential, but it’s still damned fast.)

            • ? y=x^2 increases/decreases slower than the increase/decrease in x near zero.

              Like if x goes from 1/4 to 1/2 (so a delta of 1/4), y goes from 1/16 to 1/4 (a delta of 3/16<1/4).
              The change in x from 1/10 to 1/100 is a delta of 9/100, but the change in y^2 is a delta of 0.99/100, about 10 times smaller.

              I can't quite tell, but I think you're thinking of (1/x)^2, which increases very rapidly as x approaches zero) and decreases slowly as x grows and is bigger than 1. Or you’re talking about the change in the log (from the percents?)

              But yeah, its not easy stuff to tangle with.

        • Origami Isopod

          It’s critical that we prepare the current generation of children for the modern world, so they don’t get stuck in the same low skill trap as many of their parents.

          That’s not a skill problem. That’s a political problem.

      • Well I suppose they could retrain as navigators or flight engineers. Oh wait, we don’t have those any more.

        Maybe bank tellers? Nope, ATMs got most of those.

        Oh well, they can always pump gas. Oh wait…..

      • Philip

        Software also mostly pays well because there aren’t many of us (relatively speaking) (yet). Companies constantly going on about the “skills gap” really just want the government to subsidize increasing their labor supply so they can slash our wages along with everyone else’s.

        • so-in-so

          Even in software, better keep the skills current. Other than a blip leading up to 2000, the demand for COBOL programmers is surely not robust.

    • Unemployed_Northeastern

      A fourth strategy is to tax corporations on their robots to an extent where keeping human employees is more profitable. But that would require both parties to stop flagellating themselves before corporations and offering to cut their effective tax rates to ridiculously low rates.

      • sanjait

        American industry would howl that such a tax would be a hit to global competitiveness and self-defeating in its aims, and they’d probably be right.

      • Yestobesure

        Bill Gates suggested this. I don’t think it makes sense, especially since we’re not always talking robots per se, but automation and software.
        I prefer the converse – lower the ancillary costs of hiring people. for example, universal health care so you can kill the employer mandate.

        • Unemployed_Northeastern

          Why not just advocate for UBI so as to absolve employers from having to pay their workforces at all?

    • Linnaeus

      A third strategy, rarely discussed, would be to enact policies to promote migration within the United States, as the geographic distribution of labor demand is very uneven.

      I’ve seen this discussed more in recent years, though maybe not yet enough to move out of the “rarely’ category. Thing is, at least IMHO, to do migration policies right, they have to be pretty comprehensive. Not just, say, a few hundred bucks for a U-Haul, but support for those who are moving and need to find work, support for any retraining or reeducation needed, housing support, etc. Migrating, for a lot of people, means leaving informal systems of support that can’t be replicated quickly elsewhere.

      • guthrie

        Yup. People, often upper middle class or upper class, assume that people can and should just drop all social ties, relatives, friends, etc, and pony up large amounts of money to travel across the country.

        Why not, they do it? Except that they get paid enough to actually afford to do so. And often have the opportunity to move back. Most people don’t have the resources, and most people actually want to live in stable communities with friends and relatives.

      • sanjait

        Yeah, I threw that in casually, but you’re right. Moving is not easy when you ain’t rich.

        An easy first place to start though would be to abolish the mortgage interest deduction. It inflates housing values and gets people over-leveraged with debt. Both of those things limit mobility significantly.

    • addicted4444

      Warehouse employment went up, but much better retail employment is going/will go down by a much greater number.

      Storefronts have already started closing. It’s now only a matter of time.

  • MariedeGournay

    I guess Oscar Wilde would be happy that we were finally making slaves of our machines.

  • The reality is that robots won’t replace humans as long as humans are cheaper. But once cheap, semi-intelligent machines hit the street, all bets are off.

  • raypc800

    There will be a need for repair of these robots both physical type and software updates. In WA state we are working hard to include programs in our schools that will not just be academics based. They provide magnet programs in Health Industry, Law studies, Apprenticeship programs in numerous construction positions and agriculture as well.
    There is plenty of infrastructure work desperately needed done in this country. The work is there but no funding is passed for it to be done, thus no jobs.
    I hear people whine and cry about how Democratic party members do not get things done and that is true they do not, why? One reason is people do not give them enough support via voting in true major majorities in both the House and Senate. Those that compare to FDR look at the Congress make up he had to work with and realize with the makeup of Congress today and years early do not compare to FDR Congress. Thus if FDR had this Congressional Congress to work with he would not of gotten passed the programs that he wanted.

    • Yes, there will always be shit jobs out there. Careers, though, will no longer exist for most people.

      • raypc800

        There are careers out there they are different kinds of careers ones that call for people to overcome, adapt or improvise to. Reeducation has always occurred in the work force, take newspaper industry use to be slow typesetters by hand now computers. Those that typeset either adjusted or did not life is always full of change. People just need to learn to get over the fear of something different.

        • Aaron Morrow

          “People just need to learn to get over the fear of something different.”

          The Green Lantern theory of full employment.

          • raypc800

            Actual it is called a fact of life through observation of being alive long enough to see numerous changes and public reaction to change and the unknown.

            • Aaron Morrow

              So the New Deal isn’t a fact?

              • raypc800

                I suggest you reread the posts I have made. You seem to have an error in your understanding of what I posted about FDR. Try reading the full post it would help.

              • raypc800

                Here ya go for all those that do not like to look things up http://history.house.gov/Congressional-Overview/Profiles/74th/

                Party Divisions:*

                322 Democrats
                103 Republicans
                7 Progressives
                3 Farmer-Labor

                This is why FDR was able to accomplish all that he did. Check this Congress with current one and those of the last 10 years the numbers matter.

        • BigHank53

          overcome, adapt or improvise

          Three things it’s notably difficult to do with your mortgage payment.

        • About the only career that will be left when all the jobs are gone is a career in organized crime.

          • raypc800

            Sorry Trump Administration and the GOP have that area sowed up.

        • Origami Isopod

          Old Economy Steve, ‘zat you?

    • Murc

      All of you that complain about the Democratic party answer this. What policies do the GOP offer that you support?

      What’s this got to do with anything?

      I support almost nothing the GOP does. In what way does this disqualify me from bitching about Democrats?

      • raypc800

        As my Dad always said “Bitch, whine, complain put it all in one hand, then take a crap and hold it in the other and see what carries more weight.
        Look for solutions to make things better then push it with the Dem reps in ur state or main party. Want change then act.

        • Murc

          Look for solutions to make things better then push it with the Dem reps in ur state or main party.

          Pushing those things requires complaining to, and/or threatening, them!

    • Thomas W

      “There will be a need for repair of these robots both physical type and software updates.”

      If it required a ton of jobs to repair the robots, then they wouldn’t be introducing robots.

      • BigHank53

        There are machine shops operating right now without the lights on, because none of the machines need them.

  • The bad news is your job has been replaced by a robot.
    The good news is we’re hiring robot repair technicians.
    The bad news is we’re designing robots that can repair robots.
    We have no more good news.

  • Kevin

    My mother worked for many years in a factory, recently retired (4 years ago). I was always amazed that the job still existed (was a bottle factory). It was hard work with irregular hours, but she did it for decades, and it did help our family stay “middle class”. Those jobs are going. Where do the workers go? Where do future women like my mother, people who didn’t graduate high school, go? It’s a real problem as Erik says, and no one in power talks about it.

  • SatanicPanic

    It would be great if the government could employ more people. It’s not like there’s not a ton of things that need fixing. Maybe employ people to build more low income housing.

    • sanjait

      This is a medium term solution … but a good one.

      • SatanicPanic

        Two birds with one stone! Unfortunately the political will is going to be hard to find.

        • sanjait

          Maybe not.

          ARRA was a “put people to work” bill. The American Jobs Act was another.

          Hillary proposed $275b in infrastructure specific spending and $1.8T in total new spending. Bernie proposed $1T in infrastructure specific spending. The current Dem leadership has picked up on that $1T proposal.

          It seems political support for such spending is strong and maybe even growing.

          But … we’d need to take over Congress first. So there’s that challenge.

    • Unemployed_Northeastern

      How dare you use my tax dollars to pay people to work, this is rank Commumarxicialism, people who don’t have private sector jobs should starve, etc., etc.

      – the modern right

      • SatanicPanic

        I guess the only answer is to build the wall

        • Unemployed_Northeastern

          One of most amusing and consistent lines of thought among right-wing Internet commeters is that since they pay maybe $15k or $20k (if that) in federal income taxes, they should get line item vetos over everything in the federal budget. Because self-aggrandizement.

    • Gareth

      What if you can build more low income housing by using robots instead?

      • SatanicPanic

        With innovations in pre-fab building I don’t see why not!

        • so-in-so

          The Chinese claim to have “3D Printed” buildings…

  • Unemployed_Northeastern

    “Throwing nearly a million people out of work sounds pretty great! Hard to see any down side. Democrats should just offer some tax credits for employers to train workers.”

    I hear that owning a law degree gives you a million dollars; if those million out of work people all got law degrees, that's, erm, a trillion dollars!

  • Philip

    If anyone was still wondering whether “neoliberalism” actually exists, I direct you to the comments in this thread proclaiming with absolute faith that God^H^H^Hthe invisible hand of the market will replace jobs lost to automation

    • sanjait

      Who said that? I haven’t seen every new comment but have been following comments on this post all morning and haven’t seen it.

  • Bruce Vail

    No time to comment today. Since I found out my 7x great grandfather was a veteran of King Phillip’s War I have been very busy researching military operations in western Rhode Island.

    • Murc


      • Bruce Vail

        I’m only half joking. My brother dug up an old genealogy monograph earlier this year and it turns out that Samuel Vail (1652-1695) actually was a veteran of King Philip’s War. Googling around I found out the great historian/writer Jill Lepore actually won the Bancroft Prize for her book on the war. I’m dying to read it…

  • Ross Hershberger

    Yeah, that’s me. I used to work at Saks, and folded enough shirts for a lifetime. No robot will ever straighten a table of sweaters on Christmas Eve.
    I now work in automation (laser welding machines) and I see fewer jobs all the time.
    The problem with welding and labor isn’t that a robot can do the job better than a human. It’s that a human can’t do it at all. Place 45 welds in 20 seconds with 100 micron accuracy and 1% penetration tolerance? that’s what modern product design standards require. Only a machine can do it.

  • I figure at some point the whole thing implodes due to lack of demand.

    Sure, the handful of super-rich will still demand luxury goods but we’re not talking that many people.

    People without paychecks can’t buy anything. If nobody can buy anything then there’s no point in making stuff, regardless of how efficient your robots are.

    “To date no robot has purchased a Ford” – Walter Reuther

    • Gepap

      This is why I think Marx was declared dead too early. He was correct in pointing out the massive productive capability of the capitalist system, and how it could go on and on increasing production and total wealth. He was also correct in pointing out that the internal logic of the system eventually leads to its collapse.

      You can find a way for collective ownership of production, you could find a way to create some twisted form of neo-feudalism where people have access to the surplus by surrendering power to an owner in exchange for access, or you find another way to distribute the surplus by “monetizing” something else.

      • reattmore

        This is why I think Marx was declared dead too early.

        Well, by most accounts, including that of Engels, he stopped breathing on March 14, 1883. I suppose that it is possible he survived that, but . . .

    • so-in-so

      In the 1970’s Science Fiction author Michael Moorcock wrote the “Dancers at the End of Time” stories. The main characters were a handful of remaining people who could create anything by basically wishing it (machines they knew nothing of interpreted their wishes and drew power from the universe at large to make it happen, hastening the destruction of the universe in the process). Mostly they staged extravagant continent spanning parties to fight off boredom. Might have been prescient.

  • BobbyK

    If nobody has a job who is going to buy all this “stuff” the robots are making?

    • liberalrob

      Nobody. The demand will disappear, at which point the robots will cease making that stuff and start making other stuff.

    • The only reason capital (the rich) pays labor (people) to make stuff is because they can’t make it themselves. As soon as the robots can make everything, they won’t need people any more to make or even buy stuff. The robots will just make whatever the rich need to survive in their gated communities. The rest of us will have to fend for ourselves.

      • guthrie

        Hmm, turns out the concept goes back a few years in it’s modern incarnation:


        “The day that robot armies become more cost-effective than human infantry
        is the day when People Power becomes obsolete…The rabble may think
        whatever they please, but the Robot Lords will have the guns.


        • so-in-so

          They’ll also need robot grave diggers. Lots and lots of them…

  • Jean-Michel

    I get that decoupling the link between work and dignity is not an easy row to hoe, but I think it’s worth bearing in mind that “work” is not identical to our modern conception of a “job,” which is bound up with the wage system and is really only a few centuries old a best, i.e. as old as capitalism itself. The development of capitalism completely upended prevailing attitudes about labor and human worth within the space of decades; no reason socialists couldn’t do the same if we don’t preemptively concede too much of the argument.

    • Bloix

      “the link between work and dignity might be too tough a row to hoe”
      A hundred years ago there was no such link. The most dignified people changed their clothes about 4 times a day and spent hours going to dinner and coming back from dinner. This stuff is all cultural and it can change quickly.

  • ASV

    Property tax on machines that replace human labor.

  • I’m trying to figure out to what extent overall employment will be affected by automation in the coming years. Previously there have been waves of automation that coincided with overall strong employment growth. Even in those cases, those in the affected industries are hurt, and the jobs that are available to them (if any) may be of lower quality. In the last twenty years automation has led to a lot of job losses in manufacturing, for example, and in some communities the alternative is working for Walmart.

  • Bloix

    “previous generations’ technological advancements worked because the increased jobs they created in an expanding economy absorbed those job losses.”

    Technology eliminated millions and millions of jobs for children without creating new jobs for them. No more stable boys, laundresses, street sweepers! Did we say, oh, no! we need more jobs for 12-year olds? No, we made child labor illegal!

    For the near future anyway, we could easily deal with job loss by automation. Mandatory double-time for more than 35 hours a week of work. Mandatory 6-month paid leave for the birth of a child. Free college. Full Social Security benefits at 62 to encourage retirement. Three-month sabbaticals for all, every seven years. Medicare for all to encourage people who want to leave the employed work force and would but for health care (to stay home with kids or play in a band or start a business) to do so. The increased productivity from automation would create more than enough value to pay for all of it. We are the richest country ever seen on the face of the earth and automation will make us even richer.

    This is not a technological problem. It’s a political problem. It’s about I’ve got mine and fuck you. Anyone who says different is a liar or a chump.

  • TheBrett

    And before someone says, “Derp, Luddite, Derp,” let me remind you that
    previous generations’ technological advancements worked because the
    increased jobs they created in an expanding economy absorbed those job

    They didn’t know that at the time, or what kinds of jobs might come up from this to replace lost ones. That’s why the rhetoric never really changed every time we went through a period of anxiety over technology and unemployment.

    In the mean-time, we’ve created plenty of jobs to deal with those through automation. Unemployment is the lowest it’s been since the start of the Great Recession, and while Labor Force Participation hasn’t risen as quickly (it’s been stubbornly around 62-63% for the past couple of years), it’s also not falling in a way that would suggest it’s because of anything but an economy not yet at full employment (and a country still aging).

  • Yosemite Semite

    Today, we don’t create jobs to replace those lost through automation. They are just totally lost jobs, except for robot designers. Evidence, pls.

  • Brien Jackson

    I don’t understand how you can think UBI would be too heavy of a lift but imagine anyone leaning right of center is goimg to get behind a job guarantee. The American obsession with work has certainly never extended to givong a fuck about people who couldn’t find one.

  • philadelphialawyer

    What seems to be missing here is any consideration of the lack of demand. Who is going to buy all those toys and other stuff that the robots will put in their boxes if nobody has a job and there is no UBI, or something like it, either?

  • Yosemite Semite

    Today, we don’t create jobs to replace those lost through automation. They are just totally lost jobs, except for robot designers. Evidence. pls.

    • Yosemite Semite

      If there were totally lost jobs, GDP per capita would be falling. If you look at GDP per capita in constant dollars — that is, normalized for population changes and inflation — you don’t see that. You see a constant increase. Here’s the data from FRED (St. Louis Federal Reserve) — GDP per capita in 2009 dollars from 1947 to 2017:


      You do see some drops, at times of recession — a big drop in 2009-10.

      This does not account for distribution of that GDP. If more and more goes to the top 1% and less and less to bottom 50%, then the economic situation of the bottom 50% is worse. That doesn’t mean, however, that there was a net loss of jobs.

      As check on that hypothesis, we can look at the employment figures. FRED has them from 1939 to 2017:


      Again it’s a steady rise; no loss of numbers of jobs. (The graph is spiky because the figures are not seasonally adjusted. There is a temporary loss at every time there is a recession; a big drop 2009-10.) Again, this doesn’t address the distribution jobs — what sectors declined or disappeared, and what sectors augmented or appeared de novo.

      It doesn’t do your credibility any good if you make unfounded assertions, and doesn’t help you make your case. It is surprising, coming from an economist, or economic historian. If it keeps up, soon you will have established credentials to be able to match the Republican economists.

      • Yosemite Semite

        The other check on those numbers is labor productivity. If jobs are disappearing, and the economy is producing the same amount or more (i.e., GDP constant or rising), then labor productivity must be rising: Now fewer workers are producing the same amount (or more). The numbers from FRED show that is not the case; in fact, after 2010, the number show that productivity fell.

It is main inner container footer text