What is the capitalist answer to automation?

  • While we are not currently at total automation, we are inching closer. An Oxford Study from 2013 indicates by 2033 upwards of 45% of our job force could be automated. This automation, causing a reduced need for work (thus wages) might cause (over time) capitalism to die. While there may be other possible reasons/ways capitalism may collapse, my main objective is to seek how capitalism could save itself from the reduction work/wages resulting from automation, assuming it's possible.

    Capitalism (goods are owned by private individuals/businesses). People/workers buy those goods. Business automate functions to compete resulting in a better bottom line with less labor. Less labor, less money to buy goods... and so on. Eventually reaching a tipping point of little/no people being able to buy said goods.

    If possible, how would capitalism prevent us from reaching that tipping point?


    I would cite more, but limited to two links.

    I upvoted your question because it's intriguing. I don't understand how automation could kill capitalism. Wouldn't automation just be the latest incarnation of capitalism? If capitalism dies, what do we call its replacement? Automation is a very scary thing, though. I think it's going to rank with population growth and climate change as one of the mega issues of the (near) future.

    I just followed your link and found the answer to one question - "postcapitalism." However, I don't agree with the author that we're headed towards Utopia. I think it's just the opposite.

    Isn't this the exact same thing that happened with the industrial revolution that arguably launched capitalism? (I'll give you a hint, that's rhetorical and the answer is yes.) The dominant method of producing physical goods at the time ("cottage industry") was mostly replaced by assembly lines and a great deal of automation. Why do you think this time is different? What makes you think this "tipping point" will happen this time, when it didn't last time?

    How would people stop being able to buy those goods if the increased productivity keeps making them cheaper? Who forces the people to participate in an economy that doesn't benefit them? And how could those capitalists stay in business if people couldn't buy what they produce? Who owns the factories and robots? If there is any scarcity at all, people will have a means of employment. If there isn't, *that's not a question for capitalism*. Capitalism is an economic theory, and economics concerns itself with the distribution of scarce resources and nothing more.

    The service economy :-) There are all sorts of things which people are willing to pay to have done which aren't the simple assembly-line production of goods.

    Just to make this clear - we are talking about GPAI. Such an entity is capable of putting objects and events into global perspective. When you say "automation", you actually mean "the end of all non-creative human labor". The consequences for the labor force will be fundamental. Everything, from education, to 40-hour work week will have to be reevaluated. Capitalism will have to change.

    My late grandfather (who lived, and farmed, through the changeover from horse power to the IC engine) had a saying that answers the question well. "Every new labor-saving device that's meant to replace manual work needs six men and a boy [replace by "trainee" or "apprentice" if you are politically correct] to keep it working properly".

    @Luaan Besides tech products, whose price drops as they become obsolete, and entertainment products, whose price drops as the novelty wears off, when was the last time you saw an actual reduction of price? The current trend is for a company to reduce their manufacturing prices and keep the extra profits for themselves. The market will probably correct eventually - probably through inflation - but there's gonna be some sucky times first.

    What's wrong with people simply working less? If goods are half as expensive to make I only need to work 20 hours a week to afford twice as many goods. (Then if my employer wants the same amount of person-hours as before, he can now employ twice as many people)

    @Vlad: It's not simply a matter of artificial intelligence, regardless of whether the AI is general purpose or not. To do any sort of non-repetitative task - let's say gardening, for a concrete example - that AI has to be housed in a body that's capable of many diverse movements. Even if such a body could be built, it would have to compete with flexible & self-repairing human bodies.

    @Walt You're seeing the results of continuous monetary inflation. In other words, countries are printing money and pretending it's worth the same. Almost everything is getting cheaper all the time (with or without a quality drop). The thing that prevents companies from increasing their margins *is other companies*. That's the whole point of free markets. In a real free market, you would see deflation over the last hundred years, not the unending inflation caused by politicians still using keynesian economics to steal our money :)

    @immibis We collectively stopped rewarding increased productivity with reduced working hours a couple of decades ago. Why, that is an interesting question with an answer involving politics and economics.

    "Isn't this the exact same thing that happened with the industrial revolution that arguably launched capitalism? " -- It logically can't be, since capitalism already exists. "What makes you think this "tipping point" will happen this time, when it didn't last time? " -- Because the circumstances are very different. The industrial revolution produced huge numbers of jobs that shifted the population from rural to urban. Obviously the current automation wave doesn't do that.

    "had a saying that answers the question well" -- grandfather's sayings don't actually answer questions well.

    @gerrit *Someone* was rewarded with reduced working hours, but it's not the folks whose productivity has increased.

    @JimBalter Indeed; society decided to organise solidarity among different kinds of workers and permit *everybody* a shorter working week, regardless of whether or not they worked in a sector where automation was reducing the need for manual labour. In the past couple of decades society (at least in NW Europe) has become far more neoliberal (social-democratic parties first abandoned their programme, and then their voters abandoned them) and now only a small minority calls for such solidarity to return.

    @gerrit I was thinking of Paris Hilton and Eric Trump. I'll have to think more about your more sophisticated argument.

    This question is spot on. People are thinking of automation too narrowly. What happens when every single skilled job can be automated? What happens when there are no jobs for doctors, for programmers, for engineers, because artificial intelligence and robotics can do ALL of that better than a human could?

    @Kik: I think first you have to ask whether it's reasonable for AI to think about everything better than humans. From what I've seen (and I work in an associated field) that's a pipe dream. Second, as I said above, even if you do develop such an AI, you have to provide it with some way to interact with the physical world, which for things more demanding than assembly lines, is neither easy nor cheap.

    I believe the answer is simply shifting efforts into the tertiary sector. More and more people will try to convince others to buy stuff to get provisions.

    The usual answer from capitalist business owners to people who lose their jobs due to automation has traditionally been "Welp, sucks to be you, but that's progress."

    You might find this short story interesting. http://marshallbrain.com/manna1.htm

    @immibis people naturally want to earn more money than their current job allows for (up to a certain limit) and even if goods are free all of a sudden, you still have luxury goods like real estate, private jets, space tourism, etc, that would be based on how much money you have.

  • In 1800, more than 90% of everyone were farmers. Modernly in the United States, which is a net exporter of food, less than 5% of everyone are employed on farms. That's a reduction of 85%, much higher than 45%. Far from causing the end of capitalism, it launched the industrial age.

    In short, the capitalist answer is that there is always something else that people could be doing. Police departments could hire more police. Hospitals could hire more nurses. Automation leads to higher wages which leads to more consumption of other things: maid services; landscaping; construction; other things that don't have names because we haven't created them yet.

    When I was young, we had a refrigerator, oven, washer, dryer, phone, television, and several radios. We added a microwave, a computer, and a VCR. Now, that VCR is already obsolete and replaced by Blu Ray players and DVRs (or the internet). And people each have their own phone, computer, and television (which may also be the phone or computer).

    Capitalism can't tell what the future holds. But looking at the past, it can guess that the future holds something. Because our previous responses to automation has always been to find new and different things to do.

    Plus, somebody needs to think of the solutions to problems; the people that make computers faster, phones smaller but more powerful, someone needs to come up with HOW to actually do that. And there are plenty of things in this world we still need to figure out.

    This is a good answer...as it very much *is* the capitalist's answer. (But it should be pointed out that agricultural automation has *not* been an overall good thing...)

    Also, it should be pointed out that this whole idea is going to start changing very drastically. For the past 100 years, automation was mainly a replacement for hard labor. Going forward, we're finding that automation is increasingly a replacement for light labor and white collar work as well.

    @blip: Going forward? Probably the first case of white-collar work being replaced would be _direct dialing_, the replacement of telephone operators by automated switches. That started in the 1940's. .

    Yes, we can keep producing more rather than working less. Until we reach limits.

    I find it interesting that you point to a question where you ask, Are there limits? And get two answers that say, Not as far as we know. That doesn't seem to support your thesis, that limits exist. And replacing high education work with automation could be helpful for inequality. Currently one of the problems is that automation tends to replace uneducated workers (often called blue collar workers). This makes things unequal, as demand for blue collar work drops while white collar work increases. If both are equally subject to automation, then inflation is less of an issue.

    I disagree. At some point (say in a 100 years) the automation reaches a point where it outperforms humans in every domain, even doctors, lawyers and artists. Humans will be simply obsolete by then, there won't be any job for them. The only role left for humans would be ownership of those robots & co. u̶n̶t̶i̶l̶ ̶t̶h̶e̶y̶ ̶g̶o̶ ̶f̶u̶l̶l̶ ̶s̶k̶y̶n̶e̶t̶

    @Floern Why would you expect that to happen? It's a prediction that has been made over, and over, and over again, and the result has always been the same - people found other things to do. Why do you believe "in 100 years" would actually be a point where humans are unemployable? If you make automatons that can do *everything* humans can do, you've just created a massive slave empire; not to mention that regardless of quality or cost, people seem to be turning back to "hand made" products for many reasons. Would you expect that to diminish as well?

    Personal assistant were replaced with all the technologies in a smart phone... I think a big problem is that it's very hard (need a lot of imagination) to think of what we will have in the future. There's a lot of job now that people couldn't even imagine would exists. We can't imagine all the possibility that we will have in the future (just looking at the possibility of drones is huge).

    We don't know what's going to happen. But my idea is that automation reaches a point where all those "new jobs" are immediately occupied by automation itself, because the robots are better suited to do those jobs from the very beginning. So humans will drop out of the production cycle, and I have no clue what would happen then.

    If your numbers were right, that means the US had 4.8 million people doing agricultural work in 1800, and 15 million people doing it today. So that's *still* 3x more farmers today than then. Since the USA's birth rate is now well below replacement, perhaps this "problem" will eventually take care of itself.

    I wish I could give this answer more upvotes. It should be noted that people have been raising this same concern for hundreds of years (see: Luddites.) Yet, the predicted drop in employment has never actually occurred. The prediction has been wrong every time in the past and there's no reason to believe it won't continue to be wrong going forward.

    @Brythan, sources for the 90% and 5% numbers? (Although they sound right to me) And I'd hesitate to call a 90% to 5% drop an 85% reduction. It's really an 85 point drop, but more importantly, you're talking about different bases. In 1800, the US population was 5,300,000. http://www.u-s-history.com/pages/h980.html Now it's over 300,000,000 https://en.wikipedia.org/wiki/2010_United_States_Census The farming population went from 4,770,000 to 15,000,000 (an increase of over 200%). Although the non-farming population increased many times more than that. Be careful with your percents.

    Interestingly, farming might be the industry people move back into. Industrial farming has the highest productivity per unit of human labor, but organic farming has the highest productivity per unit of land. If industrial farming cannot meet the food requirements of a growing population, the market will support a return to old fashioned, labor intensive farming methods.

    The VCR may be obsolete, but the task it does isn't. And despite TV, VCRs, Blu-Ray players, and internet video, plenty of people still read books. Even if some of them happen to read eBooks rather than (or in addition to) paper ones.

    @jamesqf Not just eBooks - don't forget audio books. It's still a book, but you don't even read it. Not my cup of tea, but certainly getting quite popular.

    @the_lotus `Personal assistant were replaced with all the technologies in a smart phone` - that's certainly not the case in any office I've worked in; personal assistants are still very much required due to the complexities of organizing an executive's professional and personal lives. There will always (the singularity notwithstanding) be the need for a professional PA to interpret the needs and wants of business people.

    @Spratty: Indeed, given sufficent money and need, some of us would likely hire a PA just to avoid having to deal with a smartphone.

    How does automation lead to higher wages? Surely not for the workers being automated out of their job...

    The human mind is more powerful than a Turing machine, the Turing machine encompasses every form of automation, therefore automation will never displace humans.

    @yters "The human mind is more powerful than a Turing machine" [citation needed]

    @Walt, A human mind proved the halting problem. A Turing machine can't do this, since it'd have to run every possible Turing machine and check whether any ran forever.

    @yters Those are two different things. A human mind *proved* that the halting problem cannot be *solved* by a Turing machine. A human mind did not *solve* the halting problem, nor has their been any proof that a Turing machine could not *prove* that the halting problem is unsolvable.

    @Walt, how else can a Turing machine prove the halting problem besides running every single Turing machine? It doesn't even seem to make sense to talk about Turing machines proving something. The very act of creating a proof appears to be unique to the human mind.

  • The answer from a capitalist's point of view is fairly straight forward. As demand for certain types of labor fall, demand for other types of labor will increase and workers will need to gain skills in other areas in order to maintain employment or for their own businesses to succeed.

    A comment to your question alludes to this. There used to be a huge buggy whip manufacturing industry when horse buggies were the standard for transport. Today that industry has (mostly) died out. Capitalists would argue this is a Good Thing™, because there is nowhere near the level of demand for as many buggy whips so producing them in large supply today would be a giant waste of time and energy. This scenario is what proponents argue is one of the major selling points of capitalism: because of the decentralized nature driving market decisions, the market itself can react much more efficiently to changes in reality than any centralized government planning office could. A government could come up with a few solutions that may or may not work, but the market itself can try millions of solutions to a demand in parallel and what works will survive.

    In terms of automation, I don't think any capitalists have all the answers (simply though because they happen to be a subset of people and I don't think any group of people have all the answers), but if they do they probably won't be sharing them publicly until after their IPO. In less abstract terms, all this means is that a different form of labor will evolve. What that is is really anyone's guess, but it's better to leave it to the market to decide rather than have a bureaucrat decide for you.

    Comments are not for extended discussion; this conversation has been moved to chat.

    > if they do they probably won't be sharing them publicly until after their IPO -- I don't see how that follows... companies don't disclose all confidential information after they IPO; they simply report financial statements following SEC requirements

    @AndreTerra I am using it more as a figure of speech here than a factual statement. If a capitalist has an idea on how to solve a problem, they probably won't share it until they can figure out how to make money off of it.

    Perhaps buggywhips have (almost) become obsolete, but (per Google: http://www.horsecouncil.org/economics/ ) US horse-related activites were a $39 billion/year industry in 2005. Other estimates go as high as $300 billion.

    @jamesqf That's partly the point, there are still plenty of people making money off of the collective horse industry, even though horses as a 'technology' have been rendered obsolete. It just isn't as prevalent in larger society as it used to, which I think is just fine with everyone.

    @JeffLambert I'm perfectly fine having less horse manure everywhere. =) Give me carbon emissions any day! =D

    @Jeff Lambert: Believe me, as a horse owner myself, I KNOW about people making money off the horse industry :-) Nor would I say that horses are obsolete. Not mainstream transportation, of course, but for things like rounding up open-range cattle they beat the heck out of mechanical alternatives.

    The idea that capitalism will naturally manage to do what's best for all seems like hopeful conflation. The imagery of the Middle Ages and the slums in the first decades of the IR don't suggest it always goes well. Take the process to the limit: one person suddenly solves all the world's production problems, such that no one need work at all anymore... he provides the world while he strips them of their remaining stuff, and then he has no more need of anyone, and they're left to die if we're envisioning strictly capitalism.

    So economic benefit !≡ people's benefit, and we're being fairly dishonest if we think some of the beneficial labor/environmental laws in place now (or for that matter things like public roads and schools) are "natural" consequences of capitalism adapting the system to fit it's needs. I'm not saying capitalism is all bad whatsoever, but thinking "it'll probably work that way because it has in the past" is dangerous and unscientific, akin to betting everything that Moore's Law would continue for the next 100 years.

    If he's providing the world, how is he stripping them of anything? Your logic is fundamentally flawed and it seems like you're trying to just find confirmation for your biases rather than trying to disprove yourself. Also, you can't discuss economics without first studying Economics, just like you don't trust a layman to offer you medical advice.

    I don't think the demand for other kinds of labor will necessarily increase merely because one kind has decreased.What will happen is that the supply of that other labor will increase, bringing prices down. I envision a future where all production is handled by oligarch owned robots and everyone else just gives each other blowjobs all day while waiting for pity money from the government. There still will be elections, but they are conflicts between oligarchs.

    @ClintEastwood If that's what people choose to do that's OK. If society moves that far to where even selling of such services is morally acceptable enough to be legal, then all you've said is just a restating of my argument. I have a relative who has a neighbor that has supported herself for decades mostly as a dog sitter, perhaps another profession that could see another big increase in employment numbers due to increasing leisure time.

    I think the biggest point in the argument that needs reiterating is that the future is nebulous and not easily predictable. No matter the competency of any single bureaucracy they'll never be able to think of the same amount of solutions as a free market. The market will definitely increase the search space also into _worse_ decisions, but a _government's_ worse decision would have a much longer lifetime since it is generally backed by law.

  • In 1930 John Maynard Keynes published the essay Economic possibilities for our grandchildren where he discussed technological unemployment ("unemployment due to our discovery of means of economising the use of labour outrunning the pace at which we can find new uses for labour"). He proposed to solve the problem by working fewer hours:

    [W]e shall endeavour to spread the bread thin on the butter-to make what work there is still to be done to be as widely shared as possible. Three-hour shifts or a fifteen-hour week may put off the problem for a great while. For three hours a day is quite enough to satisfy the old Adam in most of us!

    So say you have a factory with ten workers and you automate four jobs. There are two things you can do: fire four people, or keep everyone and reduce their working hours by 40%.

    Keynes' 1930 prediction hasn't come to pass – yet, and it may never (interesting view on why it hasn't), but it's one reasonably famous proposal to solve the problem which could still be called "capitalist" (unlike e.g. basic income and some other proposals).

    Does 40% work mean 40% the wage? That doesn't sound hopeful or even survivable in some cases. And if not why does the capitalist increase compensation or bother to automate?

    @notstoreboughtdirt Automation also means it's cheaper to produce things, so products should become cheaper as well, compensating for the lower wages. Of course, the 40% example is rather extreme; in reality change would be more gradual.

    I imagine the trickle down from improvements in wig making reducing the cost of living for wig makers is negligible.

    @notstoreboughtdirt Compensation is mostly relative, so if everyone is both working and earning 40% less then companies producing consumer goods won't have anyone to sell to unless they lower the price of those goods. It wouldn't necessarily correlate to any drop in standard of living under that scenario, the question is whether the degree of imbalance between the top/middle/bottom becomes too great, because then any sort of social mobility would go out the window since someone in the middle wouldn't possibly be able to make it to the top.

    I'd say that Keynes' prediction has absolutely come to pass! If you look at working hours e.g. in Germany, the amount of working hours per week is less than half of what it used to be 200 years ago.

    But it hasn't changed all that much since 1930 @dasdingonesin. Certainly not to levels like 10-20 hours/week.

    @dasdingonesin That's not technological unemployment, really. If you really wanted, you don't need to work all that much (provided the country you live in doesn't have mandatory work laws or fixed-rate taxes - many do to some extent, say only allowing employment at 40 or 20 hours a week etc.). It's just a trade-off between how much you value your free time, and how much you value the money you could have made if you worked instead. That's what originally drove work hours down. And how much time do people *actually* do their job out of those work hours? I know some who barely work at all ;)

    @Carpetsmoker True, but I think the reduction from 48 to 35 hours is still quite considerable, especially taking into account the post-war spike. One of the major reasons the unions fought (and are still fighting) for shorter work weeks is to save jobs while more work is being automated, which is (as I understand it), exactly what Keynes meant: Spreading the bread thin on the butter.

    Or you can shift the 4 replaced workers to help the other 6 and get a 66% increase in productivity.

    That won't necessarily work @immibis, since there may not be demand for 66% more products.

    But what if a worker wants to work more, so that he can earn more? Who would enforce these low hour work days?

    One of the things that Keynes and the like never seem to account for is that what jobs tech takes away, it replaces somewhere else. Lots of folks have brought up the buggy whip. All of those jobs gone because of the Automobile industry. Now, how many people does the Auto industry employ? The spreadsheet program was the death knell for the local bookkeeper (so it was predicted) now we have lots of Data Analysts. I work 40 hours a week at a job that *didn't even exist* 20 years ago.

    @PaulTIKI That's all logical and good, but we're fast approaching levels of automation that make large swaths of the population unemployable. Not because they wouldn't learn new things, but because computers are getting _so good_ at productive human work that even things like software engineering is starting to become more automated. What do you do when a lot of people in your society are just unable to compete with bots on _any_ level? What is left for them to do? Even the service industry which would be the fallback when nothing else is usefully left to do is becoming rapidly automated.

    @Magisch The automation is going to simply mean *more abundance* . the premise fails to account for what actually drives an economy in the first place. It gets lost in the weeds of the right now rather trying to look long term. Take a deep breath, take a step back, then take a good hard look at maslow's Heirarchy.

    @PaulTIKI What makes you think that abundance will be available to the many? Wealth is self concentrating in capitalism, so the likely scenario is that the majority will be left to starve for the scraps while the elite can now more easily then ever control them using automated technology. There is no reason to believe human nature will suddenly shift into more benevolence and less greed. What economy does there have to be if very few people can have control over every single ressource necessary for human life? At that point, the rich are just self reliant and the poor get _nothing_.

    @Magisch, you mean like in Feudalism? or the Reality in Cuban communism or soviet communism? They didn't call themselves "Rich" or "wealthy" but the top levels of the party certainly had access to a lot more resources. At any rate, look at history and you find that trying to absolutely control resources in a greedy manner is ultimately suicidal behavior. Louis XVI learned this the hard way. You are assuming that wealthy people are all like Martin Shkreli, and that just isn't true

    @PaulTIKI You are assuming that you can extrapolate from a time where humans very much still needed each other to a time where they won't. If automation indeed is so far progressed as to automate most human endeavours, there will be nothing that could possibly stop the elite. They'll be self sustaining and almost undisplaceable.

    @Magisch Economics is not a zero sum game. Also, the majority of the wealthy in the US are first generation rich. Most got there by simply living on less than they make. they tend to save money consistently over time. They don't get that way by stealing or screwing people over. There are probably some living nearby to you and you would have absolutely no idea. They also tend to be charitable givers out of proportion to most people. That is not a recipe for the kind of concentration you are talking about

    @Magisch , I am extrapolating because that is one of the best ways to predict human behaviors. Also, my prediction skills are limited because I have run out of eye of newt for other methods of divination. Seriously, Catastrophising like this never helps. Humans will *always* need each other to survive and thrive. All the automation is going to do is bump people up one level on Maslow's heirarchy

  • It doesn't have one. Full automation in a capitalist society will naturally evolve towards a dystopia without intervention from outside the economy.

    When a machine is smart enough, and can do the job of a human at lower resource consumption than the human, there is simply no reason to employ a human. So far, machines have not been smart enough, and people have been able to adapt to new jobs at a faster rate than machines can.

    Note that machines do not have to reach truly self aware or even particularly scary levels of intelligence to be able to as a group specialize in some new work opportunity faster than typical humans can (remembering that 'typical humans' are the kind of people who keep reality TV afloat). THAT will be the downfall. Each human needs to learn individually, but all robot replacements learn and benefit from each others' experience in parallel, and work at 100% from the first minute. Each human has variance in their performance, but a machine works like... well, a machine.

    Presumably we would still need humans to design the machines making at least some jobs. Unless you propose machines can design better machines than themselves and than humans can design, at which point we reach a singularity where we are defunct as a species, not just as individuals.

    This is the best answer thus far. Capitalism has no answer. The rest of the answers are ideology.

    @axsvl77 That's the best argument I've ever seen in the whole universe. Let's just stop discussing anything, since everything you don't agree with is ideology anyway ;)

    You're wrong, and not in an interesting way either. Division of labour makes sense even if I can do every single thing better than anyone else in the world - trade will still make both sides better off. Increasing productivity means you need less resources to feed the humans, and in a different way than the robots. And either the robots will be owned by humans (investment like any other), or they'll be free (if they can do anything humans can, what right do you have to enslave them?). And thinking that robots will be able to benefit from each other's experience is hopelessly naive.

    This is the only answer that doesn't discount the inevitable singularity that will occur when all work done by humans is obsolete. The day that an ai is made with human level intelligence, the next day there will be one with 10x human intelligence, and so on into eternity. A single computer mind can scale infinitely, while a single human mind can not. Of course the capitalist argument would be, that if no one has a job, then they have no money, then they can't buy things, so the people who do own things will have to sell. This doesn't work when the thing you can't afford to buy is food.

    "but all robot replacements learn and benefit from each others' experience in parallel, and work at 100% from the first minute. Each human has variance in their performance, but a machine works like... well, a machine." That's not remotely how machines work.

    @Vality I think that is an inevitability, in which case we become the horses in the parable of the buggy whip, The US horse population peaked in 1915 and is now a bit more than 10% of its peak. We're sort of a meat-based bootloader for a more sophisticated type of intelligence.

    @Luaan division of labour only makes sense when you cannot simply duplicate the more skilled labourer as many times as you want. This is why we no longer use horses for hauling goods or transport except in ultra niche cases or for the novelty. If cars and trucks were in limited supply we would still use them, but they aren't so we don't.

    @NPSF3000 That's *exactly* how modern learning machines works. Every Tesla with auto-pilot reports its "driving experience" back to the central servers, and makes all auto-pilot cars that much better. Every new Tesla is sold with the latest data, not needing to drive for a while to learn and improve. *That's* the robots' edge.

    @StevenArmstrong if that's how machine learning works 'work at 100% from the first minute' how come self driving technology STILL hasn't managed to become truly autonomous?

    @StevenArmstrong Horses are *much* more expensive than trucks and cars (though even then, they keep their niche - e.g. forestry, low-level infrastructure, sport, art...); but that's not really the point here. Horses aren't workers, they are tools. They are force multipliers, not actors. Until robots get to the point of independent action, they are tools - afterwards, they are *people*. Slavery has mostly been forbidden for a while now :)

    @Luaan Workers are absolutely tools. Labour is just another exploitable resource. Horses and people are only distinguished in the eyes of corporations by the laws guarding them, and laws are forces *outside* the economy, like I qualified at the start. What do you think is going to happen when *people* are much more expensive than *robots*? We do not require independent action for this transition, as the lower 50th percentile of human workers aren't clever independent thinkers who are constantly solving novel problems. They're mostly just running off a script computers can't yet follow.

    @NPSF3000 I think you're possibly being a bit obtuse. Everyone new to a field starts off from scratch. To hire, you have to find a candidate, then test qualifications before hiring, then there's always training costs afterwards to catch them up with specifics to your company. Once *at least one* robot is qualified, install the right software and every subsequent robot is good to go from the start. No headhunting, interviews, or weeks+ of poor work while they get up to speed. Same performance as the last one on day 1. People also don't come with warranties, nor can you just sell them later.

    @StevenArmstrong " Once at least one robot is qualified, install the right software and every subsequent robot is good to go from the start. " Sure. As soon as a car is great at self-driving in SF it'll be fine in LA, or Dallas, or Shanghai or Berlin or London or outback Aus etc. And, once it works in one kind of vehicle (in sunny conditions) it'll work in all other kinds of vehicles in other conditions. It's obtuse to think that some hand-wavey simplification of technology is remotely an accurate depiction of how complex systems are developed or work.

    @StevenArmstrong Have you ever watched, say, a worker on a construction site? They don't need to be "clever independent thinkers" to be hard to replace. Robots are very specialised, and will be for a long time, because flexibility is absurdly expensive; but I digress. You still haven't answered the main question - how exactly do people become more expensive than robots (ignoring unions, of course)? What advances in productivity and efficiency would make robots cheaper without a comparable discount for human workers? How many people can a single farmer feed today?

    @StevenArmstrong And even if that somehow happened, what prevents the "lower 50th percentile" from having their own economy, independent of the "robot world"? Are you going to force them by violence? When would the wealth and power sit in such a society? How do you become rich if you have noöne to sell to?

    @Luaan What prevents someone from someone with automation from entering their economy? Are you going to hold them out with threat of violence? That's unlikely considering the side with automation could just drop a celestial body on you. Wealth would be largely obsolete by that point, replaced by control of resources and raw power - whoever had the largest automated base would hold the most power. There would be no need for exchange of anything except information. And please remember I said that there were no solutions that could arise from *inside* the economy. Outside is another story.

    @Luaan What is wealth besides abundance of ressources? If 1% control the majority of the ressources, they don't have to sell to anyone to be wealthy. And they certainly have the means of supressing the 99% then...

    @Magisch They have to trade with someone to *keep* being wealthy - since more wealth is continually being produced, they'd just become poor over time. That was one of the things that destroyed old-school aristocracy - they couldn't cope with losing their near monopsony on employment. One thing people tend to widely misunderstand is how much of a mobility in wealth there is. The people who were poor twenty years ago aren't necessarily poor now. And you're still assuming they're controlling all the resources they care about - as if owning all oil had any bearing on having pepper on the table.

    @Luaan That's a fallacy. Ressources that equal wealth have a hard cap, dependant on ecosystem. Capitalism isn't a constantly producing game, with sufficient efficiency and automation it becomes a hard zero sum game.

    @Magisch "Resources that equal wealth?" Hardly. We're producing lots of wealth in things that aren't really material. Heck, I do that for a living :) Do you really think that a piece of software has no value just because it doesn't require "raw resources"? Do you really think your imaginary 1%er will just sit on his wealth doing nothing? He's still going to have needs he wants fulfilled. He's still going to try and make his wealth bigger and provide for his family. How exactly would your zero-sum scenario work? Where's the suffering customer who can no longer provide for himself?

    @Luaan Human survival is precluded on ressources. Everything else is extra and thus irrelevant to this discussion.

    @Magisch Maybe in a communist utopia. Real humans have a lot more needs than "survival and reproduction" :)

    @Luaan It's how humanity evolved and has lived for the better part of 200.000 years.

    @Magisch We know very little about what humans 200 kya were doing, but it's a good bet they did a lot more than bare survival. Humans are adaptation executers, not utility optimisers - even though the only goal of our evolution is to make more humans, humans themselves actually care about things like fun, art etc. Survival is just the most basic need, it certainly isn't the only need.

  • I'm not sure if Tim Burton is a Capitalist or not, but he gave a really good illustration of what tends to happen over time in his version of Charlie and the Chocolate Factory

    • Charlie's father works capping toothpaste tubes
    • The factory buys a machine to cap the tubes. Charlie's father is fired
    • At the end of the movie, Charlie's father is the man who repairs the machine that replaced him

    Capitalism's chief feature is that it self-levels. For instance, if you produce a good that is too expensive, nobody will buy it. Henry Ford realized that and began to produce cars at scale (which drove the cost down) and paid his workers enough so they could afford to buy the product they make. Today, perhaps half of every car is produced by automation and Ford still sells lots of cars.

    Remember, the goal is consumption of goods and services, not creating jobs (which is purely a political goal). I found this comment to be insightful

    Luddites have constantly engaged in the fallacy of looking at jobs as an end in itself, rather than facilitating consumption as the real end. Production is merely a means to the end of consumption and the real objective is to produce the most goods and services with the minimum effort. This fallacy becomes extremely apparent if you consider a simple case of a single person on an island. Obviously his objective is to build himself a nice house, grow himself enough food, build enough nice things for himself etc. with minimum effort. His goal is most certainly not to work 40 or 60 hours a week irrespective of what that labor produces. His goal is to produce the maximum set of things that he wants or needs with the minimum effort. He would be overjoyed if robots did 97.5% of his work needing him to work only 1 hour a week. Nothing fundamentally changes when multiple people are involved who do a relatively more complex form of barter using a money system to trade with one another and produce those set of items that they enjoy a comparative advantage in producing and trade with others to get access to other items that they have no comparative advantage in producing.

    About 70% of the US was engaged in agriculture in the 18th century and luddites always feared automation in agriculture resulting in loss of jobs. Today about 2% of the US is engaged in agriculture since the average agriculture worker has his productivity greatly enhanced by technology, and the remaining human capital has been freed to engage in other productive endeavors.

    Robots replacing human jobs will have the exact same effect as what technology has had so far when it destroyed human jobs, which is improve overall human productivity, leading to higher real incomes and greater prosperity.

    Thank you very much for sharing that comment. It describes my opinion about this topic better than I ever could, and it was driving me insane not having a way to express it.

    The Henry Ford thing is a myth. He didn't pay his workers high wages so they could buy his products (that's just as ridiculous as feeding dead humans to humans as food to produce electricity). He paid his workers high wages *because nobody wanted to work there*. People simply didn't want to stand next to a belt the whole day doing one thing over and over again. The worker turnover was horrible, so he had to keep increasing wages to increase his profit margins. You only make a trade if the value you are getting is higher than what you're giving away, and his workers simply didn't feel that way.

    The island analogy falls down because it ignores distribution of wealth. Say you had an island with 10 people, and it took 40 hours / week to build and maintain housing, and 40 hours per week to farm the available land. You could have one full time builder, who provided a home for a full time farmer in exchange for food. What have the other 8 people got to trade for the food and housing they require?

    @thelem I think you're ignoring something important, though: **scalability**. There are many economic systems that can be made to work on a small scale (like 10 people), but not on a large scale (300M people).

    @Machavity I don't understand. My argument remains the same if there are 300M people on the island, with 10% working in housing, 10% in farming and 80% unemployed.

    @thelem No, it falls apart at 300M. With 10 people the goal is mere survival. With 300M, the goals are far more diverse than just farming and housing (I can't think of any economy with only food and housing as their sole goals). You'll never reach 80% unemployment because there will always be demand for labor to reach the diverse goals of others. Even in poorer countries like, say, Mexico, the unemployment rate is under 10% consistantly

    @Machavity That's the argument made in other posts (e.g. by Brythan). I'm just trying to show why the one person island analogy doesn't work (because it ignores scalability). It suggested everyone could work one hour each per week, but if a community only needs to work 1 hour per person per week on average, what is to stop one person doing 40 hours of work, and leaving 39 with nothing to do.

    @thelem "What have the other 8 people got to trade for the food and housing they require? " If they can't trade anything of value to the carpenter or the farmer for housing or food... they start trading among themselves. There's nothing preventing any of the 8 from realizing there is demand and starting to fill it.

    @Machavity There will, eventually, be near 100% unemployment. Human beings are not efficient, they just seem like it so far because we have nothing of comparable intelligence to compete with. When we do have an intellectual competitor, we will either have to augment ourselves to the point of not caring about employment anyway, or we will be unemployable. Either way, no more people working.

    I think the story of Charlie and the Chocolate Factory is better credited to Roald Dahl than to Tim Burton.

    @Evargalo Partially. Burton invented the part where Charlie's dad gets a new job repairing the machine.

  • Capitalism doesn't have an answer, and doesn't need an answer. What people always forget when they worry about automation is that prices for things get less. At the turn of the century, people would pay 43% of their income on food, just to stay alive! That has gone down to about 5% now due to automation! (I can't find a source chart for before 1920's.)

    us spending on food expenditures

    What will happen when people can furnish their entire home by visiting the dollar store? (They can already do that.) They will have more money left over for other things: entertainment, sports, services, creativity, travel, etc. Maybe people will only work 1 day a week to survive? Maybe they will work 5 and live like royalty did in the nineteenth century?

    Correct. Under capitalism, there is no question of automation. Instead, automation is an answer to the question of not merely staying alive but one of thriving.

    It seems to me your thesis is based around 1 industry, food. There are other necessities that are more expensive than in the past with little gains in income. Not sure i fully understand your answer.

    @jharris8567 http://www.huffingtonpost.com/2014/09/08/3d-printed-houses_n_5773408.html _Everything_ gets cheaper, in _all_ industries.

    buying enough *healthy, organic* food for 5%-10% of your income? Not sure.

    Do note that if you're living in the first world you're probably *already* enjoying a higher standard of living than most 19th century royalty while working only 40 hours a week. The king had more human servants, sure. But the food was worse, the healthcare was worse, the climate control was worse, the indoor plumbing was worse, the travel was worse, and most of those servants went to doing the stuff that Alexa will do for you today for free.

  • While labor is marginally valuable there is no problem. Fewer apple pickers needed means we've freed up people to be beekeepers. Fewer beekeepers needed means we've freed up people to be carpenters. And so on.

    The imagined danger is when labor is not marginally valuable. When there is nothing productive a person with a free day could do that would cover the cost of living a day. A capitalist might laugh at this possibility and offer you a job at a not quite competitive wage, proving pretty clearly that day is not today.

    And thinking a ahead on a whole system level is not expected of a capitalist, but if they did they might say something like: If a person alone can't earn his daily bread how was food found to feed him to adulthood? We will have reached the carrying capacity of the system and population growth ahead of resource growth is expected to be bad. Let him look for charity.

    "And thinking a ahead on a whole system level is not expected of a capitalist" = interesting point!

    " thinking ahead on a whole system level is not expected of a capitalist". That's in fact the core point of capitalism. Since the whole system includes all personal preferences of millions of people, you _cannot_ think ahead. Capitalism is known to be sub-optimal with respect to the millions of preferences, but it is _robust_ in regards to getting those preferences slightly wrong.

    @MSalters Depends on what you call "optimal". If you mean "satisfying all needs and preferences fully", sure. The only "optimal" solution would be to erase people's preferences (which, granted, *has* been attempted with indoctrination) :P But as for "we know of a better way of simultaneously fulfilling all those preferences as much as possible"... I'd like to hear of such an idea. Not to mention that capitalism keeps the responsibility on the individuals - they get to choose what they value more, to the best of their knowledge.

    @Luaan: I think we can agree an optimal solution is at least Pareto efficient. I can make the stronger statement that capitalism isn't even Pareto efficient. That is not a _constructive proof_ of alternate systems, though. That's the problem with an infinite set of alternatives. In a finite set of alternatives, I can find the best solution by comparing them, and the notion of constructive proofs is not that important.

    @MSalters isn't capitalism Pareto efficient if you make the common (unrealistic) simplifying assumptions, like perfect information and no barriers to trade?

    @blip "interesting point!" more like unsubstantiated claim

    @AndreTerra there seems more evidence of it than not.

    @notstoreboughtdirt: A Pareto inefficiency in capitalism means that two parties can make a mutually beneficial trade, but don't. The reasons you mention can hinder such a trade, and therefore cause these inefficiencies. IIRC there are also numerical issues, such as non-existent derivatives of demand curves. (you can't have 3.1415 cars).

    "more like unsubstantiated claim" -- it's hilarious to see someone offended at the assertion. This is the worst sort of kneejerk ideology.

    @blip there are plenty of examples of "capitalists" (and let's pretend for a minute that is an actually useful word to describe whatever class of people the OP purports to describe) who think forward. in fact, those who do out-compete those who don't quite often. if you take any business strategy course you'll be served with numerous examples. one such example is IBM

    @AndreTerra that there are forward thinkers does not negate the quote. No one said there aren't forward thinkers. Just that, in general, in the quest for making a buck, that's hardly a requirement. (IBM is also a terrible, terrible company but that has nothing to do with this conversation...just that I've had to deal with their shitty software and shitty consultants for waaaayyyy too long. There is nothing forward-thinking about their product line. :) )

    @blip the burden of proof is on whoever made the claim, not me. my example is merely an attempt to falsify it. as for IBM, you're too focused on IBM today when I actually mean the history of IBM. it's a prominent example of incredible foresight in business strategy.

    @AndreTerra Again, pointing out the exceptions to the rule doesn't negate the rule. There's plenty of evidence to reinforce the rule...the banking debacle, the bankrupting of the US Auto Industry. The first Internet Bubble. The second internet bubble. The third...

    There is never nothing productive to do in a day. When all of the critical labor has been replaced with robots, food, clothing, and shelter will effectively be free and labor will be redirected to creative outlets- art, music, games, etc... IP law will need to be reformed in preparation for this.

  • The answer is that even when machines can do anything, they are not necessarily the best ways to do it.

    Many developing countries do manually many things that we automated away ages ago. And it makes perfect sense: in those countries, wages are so low that paying for a machine actually wouldn't get your investment back. Automation only kicks in when labor is scarce; if it isn't, and wages stay low, then humans will keep being the cheapest machine.

    The problem arises when labor is locally scarce despite mounting unemployment; for instance because of unionization, minimum wages, or people not having the required skills, or monopolies being allowed to keep prices high on necessities. But that's a problem for politicians to solve.

    Without political interference, price of food and building will go down so much (thanks precisely to automation), compared to what skilled technicians and machine owners make, that humans will always be cheaper at something.

    Of course it depends if you like such a world.

  • More hair dressers!

    As the primary production industries employ fewer and fewer people, service industries grow to take up the slack.

    There will always be a market for having a real flesh-and-blood person use their time on you. It feels good, and will always be in demand.

    In the old days rich people had servants to wait upon their every need. Including needs other people simply didn't have, like help getting dressed.

    These days it is more common to go to a restaurant, hair dresser or some other place and pay the people there for their time and attention. (And, I guess, their actual work)

    Look for example at the role of the "store greeter". I think most people would agree that this person does not do anything productive. They still have a job, they still earn wages and they still buy other goods.

    Look forward to a future of more store greeters.

    I have never seen a store greeter. Where in the world do those exist?

    @gerrit I've seen them on The Simpsons!

    @gerritt - I have observed them in grocery stores in the United States. Wal-Mart and Hy-Vee may both have greeters, though they are often semi-retired older people.

    Where I live we don't have store greeters. Instead, we have automatic checkout systems. I heard in the US you get them too (Amazon Go).

    Home Depot also has greeters. They're typically older people who are there to help the weekend warriors find what they're looking for

  • A simple examination of history should put this fallacy to rest.

    Automation has been replacing manual labor for as long as manual labor to accomplish a specific task has been performed, at least several thousand years. And most of us still have jobs, despite the automation and a growing population. How did that happen, despite all the predictions of gloom and doom?

    Probably the first bit of automation was the animal drawn plow. One person with an ox drawn plow could cultivate the same amount as many people with hand tools. Did that put the many people out of work? No, it did not - now that everyone wasn't engaged in food production, some could specialize in more advanced fields, like building better plows, breaking oxen to harness, building and operating wagons to transport all that extra food, or what have you.

    Moving forward, consider the case of Ned Ludd, a weaver in late 1700's England, who smashed up some automated knitting machines. Granted, Ludd appears to have been a ne'er do well who was championed for the wrong reason, but the coming of the automated loom as opposed to hand spinning yarn did not put everyone out of work.

    The automated loom dropped the price of clothing to the point where more people bought clothes more frequently, creating more jobs at the automated factory, plus more jobs to meet the increased demand for wool and cotton, plus more jobs to transport the raw goods and finished clothing into the new stores with new jobs to sell the clothing to meet the increased demand...

    Eli Whitney automated two manual processes: extraction of cotton (the cotton gin) and a standardized firearm production line (previously, guns had been pretty much handmade). Both reduced the number of people necessary for the task, neither ended up putting people out of work. Drop the price of the item being produced, demand goes way up, production goes up, more jobs, and more jobs to supply those factories, and to transport the goods.

    Arguably, the coming of the steam railroad put a lot of horse drawn wagon operators out of a job. But, the steam railroad so boosted commerce in general that there were plenty of jobs to go around.

    And, one of the most ironic cases of automation and job loss: Henry Ford's auto production line. Ford automated much of the process of building an automobile, previously done by hand. Ford dropped the price of a car by an order of magnitude, by lowering labor costs. Did this eliminate jobs? In fact, it did the complete opposite. Ford's pay of $5/day for work on his automated assembly lines essentially created the middle class, and in the process, Ford created a lot of customers... for his cars. More cars sold meant more demand for metal (more mining/smelting jobs), more demand for gasoline (jobs drilling, refining, and selling gas), more need for repair (jobs repairing automobiles)... it just kept growing. The cheap automobile made a lot of new jobs possible that weren't efficient before, like traveling salesman, delivery person, etc...

    Finally, let's look at a contemporary situation. E-commerce. It is putting some brick and mortar stores out of business. Oh, dear, that's automation killing jobs, right? In fact, the opposite has happened. E-commerce relies on delivery, with UPS and FedEx seeing explosive growth in delivery. More jobs. The websites must be maintained - being a good web developer today is a fairly lucrative occupation. E-commerce with its lower prices and greater selection has increased sales, which increases production, which increases jobs elsewhere.

    For as long as civilization has existed, automation has been eliminating jobs. If you look at previous examples, not by today's standards, but within the context of their time, you should be able to see that for every job that automation eliminates, the results of the automation creates more than one job.

    Automation is not only not detrimental to employment, it is mandatory for increased employment.

    Well said. the panic mongers always bemoan the loss of a sector and never come to grips with the fact that another sector will take it's place, and we can't accurately predict what that would be.

License under CC-BY-SA with attribution

Content dated before 7/24/2021 11:53 AM