The Long-Term Jobs Killer Is Not China, It’s Automation
nytimes.com
nytimes.com
We currently don't have a social understanding for navigating a world where little human labor is required. The resentment of the winners (capital owners) by the losers (jobless masses) may lead to great upheaval, to be harnessed by willing politicians of short-sighted vision.
What will it mean to work and contribute to society in the future of machines?
Consider, the world before the internet when the postal services was doing all the heavy lifting for communication. Once we got internet, a huge chunk communication was replaced by email etc. But, it also gave rise to eCommerce, on demand delivery services etc. And, the same postal services companies are doing better than ever before.
Yes we do.
The industrial revolution had far greater impact on labour than any of this mysterious AI will ever have.
The first 'machines' powered by coal etc. instantly 'unemployed' hundreds of people per unit.
Go ahead and look at the productivity/capita charts for the early 19th century - UK citizens were multiple times more productive than citizens of any other nation. It was an astonishing explosion.
Imagine how many people a single train unemployed?
Or the weaving machines that did the work of hundreds.
The impact on labour was definite, direct and measurable - unlike a lot of the 'soft' impact that today's technology has (did MS Word really unemploy 'secretaries' - or was it a host of factors?)
But what happened during the industrial revolution?
Unemployment went down, wages went up. In fact the cost of human labour rose dramatically.
The surpluses from the machines went into the economy, basically providing the foundation for the modern consumer economy.
95% of folks back then worked on farms, mines - or did manual labour/service.
The industrial revolution created the middle class: lawyers, entertainers, restaurants, mass market clothing, design etc.. Electricity created massive new industries (think of what you could not do without electricity!).
The economy has been diversifying rapidly since the start of the 19th century in the UK, and everywhere else about 50 years later.
We will continue to do this.
In 1960 there were 3 channels. In the 1990's 50. Now it's 100's + online offerings.
There were no 'pro athletes' in 1930. Now we have massive industries around pro sports.
Working/middle class people did not travel much back in the day. Now they can travel around the world.
There were no consumer electronics past the radio in 1940. Now the are zillions of devices.
In the 1920's cars were made by people - now automated assembly lines.
And FYI - for the last 25 years and for at least the next little while - China (or rather, Mexico, China, India and other low-cost places) are by far and away bigger job killers than automation.
'Automation' as we understand it is not as fast a process as we think it is and the days of 'robot replaces x people' are long gone, it's a more complex and nuanced equation.
Try this one: tell me which people you know, in which industries, have outright 'lost their jobs' due to some AI or machine? Not so many. Customer service? Possibly. Some banker analysts? Possibly. But even in the later, there are other things for them to do, and the bulk of the staff on both sides have more to fear form India than Microsoft. In fact, this is happening right now all over North America: people training dudes in India to do their jobs, after they've been told they've been outsourced ... (train your replacement if you want your severance!) ...
Also - since 1900 - the size of government has expanded dramatically. I'm not 'against government' or anything, but many gov. agencies are abhorrently inefficient. I have a distant family member who's friend, a low-level thug dealer who was recently arrested. The RCMP (Can. police) spent zillions tracking, investigating this guy. Two years of surveillance, bureaucratic justice issues, lawyers - yada yada. So it's not so nice, but we find ways to 'distribute the surpluses' of the new economy in ways, even if they are not very efficient.
The trick is to 'find things for people to do' in a manner that is reasonably efficient and fair.
Check out GDP/capita during 19th century:
https://en.wikipedia.org/wiki/Great_Divergence#/media/File:B...
Were there was more automation, wealth exploded. (This chart does not address wages, but there are others that do :) )
>In 1960 there were 3 channels. In the 1990's 50. Now it's 100's + online offerings.
Exactly, we're already beyond the limits of what we can consume. There won't be jobs producing more stuff when there is no one to consume it.
No, because in 1960 everyone had the same thing.
Now - you get programming for your niche, which is different than that of the dude next door.
Take clothing: even in the year 2000 people did not really wear fashion. We had 'the Gap' and not much more. Everything else was expensive.
'Fast fashion' has changed the world. Young people can now afford an absurd variety in fashion. Nobody 50+ ever thought you needed a 'new pair of pants' every season. Now it's normal. Clothes are almost disposable. That's a consequence of increased productivity and adapting demand.
"It's all subsidized through foreigners plights and profited upon wildly"
This is highly speculative, and I don't think in any way substantiated.
"they just exploited other countries through capitalist guises instead of barbaric colonisation."
Borderline racist.
American consumers are fuelling a massive wealth creation boom in places like China and India, where 100's of millions are coming out of poverty.
Trade is generally good for both sides, and it has benefited China/India etc. massively.
It's funny what you call 'exploitation' is 100's of millions of people with massively increasing wealth.
If you want to see China and India desperately poor again, we can 'turn off globalization'. A few more Americans will have jobs, but it will be back to the dark ages for many economies.
I need someone to burst my bubble from time to time.
Also, as a side note, I for one would be delighted if the future included people brewing delicious beer for one another as an important part of the economy.
It should be added that wages going up and the emergence of a middle class was not a direct result of increased productivity. The labour movement had to fight hard to ensure that those productivity gains were shared with workers.
The problem we face with automation is that there is not much of a collective 'working class' to fight that fight.
I fully agree that the labour movement was essential.
BUT - wages did definitely rise far before labour got going - and wages could not rise as they did without underlying productivity gains.
But yes - the labour movement was essential.
Absolutely. The challenge with automation is ensuring that the increased productivity benefits all of society (not necessarily equally). As useful as capitalism is, it doesn't really have a mechanism for doing this properly. The solution will always have to be political.
I think a large part of it has to do with how disruptive computers and software are. One person can walk in and write some software that makes an employee 50x as productive. How do you compensate that? They enabled your employee to create profoundly more value. If compensation is in any way tied to the value of the work being done, the compensation should obviously be similarly profound. You have to divorce value of work and compensation from one another completely to avoid ending up paying software creators very highly and practically creating a special class for them.
As software developers, it's literally our jobs to put people out of work. My software is specifically designed reduce the need for increased labor as the company crows. In my industry, legal, entire departments (like night word processing) no longer exist.
The fact is you cannot extrapolate from 100 years of human history and say everything will continue on. People just a few hundred years before that could not have predicted anything about modern employment.
It's a lot more than 100 years. The invention of the scientific process, the printing press, the popularization of steel and before that, bronze, heck even the invention of agriculture radically changed human societies by increasing individual productivity and enabling specialization.
> People just a few hundred years before that could not have predicted anything about modern employment.
Right, which is why it's silly for us to look at the technology we have today and say "welp, that's the end of human employment."
And why it's also silly to look at the past and say "meh it worked out last time, everything will be fine".
That is either a very amusing typo or your employers are not nice people.
This is the same thing, FWIW. Opportunities to use skilled labour rely on the existence of skilled labour capable to fulfil it. The reason is that skilled labour shortages push up the price of labour, and that reduces the viability of businesses that use it as an input, and thus reduces the number of opportunities advertised.
> Why does the person putting together my will or litigating a suit for me need to be backed by an organization with an office building, expensive executives, etc?
These exists because they are currently necessary (they are the "tools" you speak of) but they're a pure business cost and constantly under pressure to not exist. Computers have already reduced this back-end cost immensely. For example, lawyers these days all use dictation software rather than having a secretary write out shorthand. If all the tools can be automated, they eventually will.
This isn't a terribly novel idea. Companies which produce phonebooks have been suing the hell out of municipalities which wish to discontinue the practice of automatically giving a phonebook to their residents for years now. And, by the way, they win those cases. Even in municipalities where they simply want to change receiving a phonebook into an opt-in service, they're suing. That their product actually now goes beyond being worthless directly into the territory of actively destroying value, requiring the munis to put out special dumpsters so that the residents can immediately throw away the useless things, doesn't seem to bother them. They want their money and they don't care how or why they get it.
There was an article about the issue with writing software which does legal things in the Communications of the ACM a year or so ago. I haven't kept up with the topic, but at that time several states had adopted the laws and they were being considered in others. Having lawyers pushing new laws is a pretty easy sell, though, as most legislators come from law in the first place. Some of the states have a nice exemption that just really hammers the intent home - you can run a software service that lets people fill forms, get generated documents, etc... if you pay a lawyer for each thing it does. They don't have to be involved in any way. They don't have to review the documents and put their name on them or anything like that. They just have to get a check. Honestly though, I think that is better than only making it legal for lawyers to use the software. I fear a future where everyone is yolked and herded into a cubicle to sit there and push a single button mindlessly, prodding the automated system to do its thing simply because society and management want people to have to do a job.
That has not been my experience at all. As tedious tasks are automated, the people who do those tedious tasks are no longer required.
Ultimately we are lowering the cost of our "product" but it's not to enable more people to afford it. We need to do that because less people can afford the service as is and we need to cut costs just to stay in business. It's a race to the bottom, not a race to the top.
That almost never happens in my experience. Deploying those people to do new things requires training them on those new things, and most companies aren't willing to make that investment, especially since said people are usually near the bottom of the totem pole and the fact that they just got made redundant by software is regarded by management as proof that they are just cogs.
The global economy is in good shape.
Western economies are stagnating - but there are 5 billion people who were not even on the economic radar just 50 years ago. They were earning basically 'nothing'.
But now they are earning 'something' and it's growing fast.
There are more than 40 countries with growth over 5% GDP growth in 2015. [1]
There are 5 billion people who need: electricity, clothes, cars, tv, entertainment, sports, medicine etc. etc..
So, no. As prices fall, and globalization expands - all those products and services become accessible to these countries.
And what happens when they get reliable electricity? Governance? Consistent internet access? Their productivity goes up and they can afford to buy more.
Viewed from a 'modern world perspective' - I would be concerned, we are going to have to 'find new jobs' for a lot of people - but from a global perspective, the picture is mostly positive.
[1] https://en.wikipedia.org/wiki/List_of_countries_by_real_GDP_...
But they are not most of the population.
We're talking about employment here.
5 billion people 'lives improving dramatically' vs. 2 billion people 'lives improving slowly' is a good equation.
That those 2 billion are 'already wealthy' and a small uptick in their wealth increases the global GDP a little bit more is and odd way of looking at it.
By that measure, we should be focusing on Gates and Zuckerbergs wealth more than anything else!
The 'global GDP' is not the point here.
Lets say youre a truck driver and suddenly you lose your job. You cant transfer your existing skill set to driving a cab for very long because guess what? Thats been automated too.
There arent any other jobs that involve sitting at a wheel avoiding bad drivers and rush hour traffic and pulling all nighters. Simultaneously the chance for that person to find work in new markets becomes low. They have to train themselves.
Things did indeed come out well eventually but that better world most definitely did not emerge serenely and rationally. It took the utter annihilation of the pre-WWI world order through decades of enormous violence. Indeed, the egalitarian socialist plan and the response of populist fascist to crush it emerged exactly because of industrialization.
There can be no doubt that displaced workers are not going to smoothly transfer to become professional athletes: a 50 year old unemployed coal miner with no social safety net is not going to peaceable become a wedding planer in a big city even if he could.
Until then, there will be increasing political disruption and radicalization as the advantaged group holds the disadvantaged down believing it's their own fault for not changing careers. And just like last time, the fighting will continue until adequate social safety nets are in place.
It would be better to honestly face the events of the past and not try to convince ourselves that an idealized smooth economic shift is how it's going to work out. But unfortunately we are just at the beginning of this and likely most people in the advantaged group will ideological despise the level of social security that will solve the problem. Indeed, in many quarters, there is a fetishization of and desire to return to that pre-WWI unconstrained economy that caused the nightmares in the 20th century. So, polarization, demagoguery, extremism and eventually violence loom for now I fear.
No - the late 19th century was relatively peaceful.
"But unfortunately we are just at the beginning"
Just at the beggining of the longest period of peace and economic prosperity in history.
Donald Trump, Brexit - this kind of 'nationalism' and 'demagoguery' is not even remotely in the league of anything in the past. Not even close.
>...is not even remotely in the league of anything in the past. Not even close...
I wish this were the case. And while Brexit is only one referendum, possibly recoverable, the Front National, AfD, PVV and Trump use rhetoric that is completely indistinguishable for Eastern and Southern European fascism of the 1920s. Some of it is verbatim quotes translated. And the economic policies of these parties will only worsen the economic well being of the electorate angering them further.
We can keep making optimistic guesses about the next score years but there is also more realistic outcome:
We do not in fact live in a magic time at the end of history. The post WWII (relative) peace is not remotely an inevitable state of affairs. It is an incredibly delicate thing maintained by moderate democratic global nurturing. Inept leaders set on dissolving alliances and treaties combined with aggressive rhetoric and random hostilities executed solely to excite a demagogue's support base could break that peace in any of a dozen powder kegs world wide faster than the bullet that killed prince Ferdinand. Indeed, if NATO weakens enough, as many on the far-right wish, and Putin starts loosing popularity, god help the Baltic states.
Babe Ruth, arguably one of the most famous athletes ever, played from 1914-1935.
Correction: there were 'few' pro athletes. Today, it's almost a career path. Back in the day, it was mostly amateurs.
> tell me which people you know, in which industries, have outright 'lost their jobs' due to some AI or machine
Right now AI is still a future technology. The current AI algorithms are the equivalent of the early GSM phones, which looked like car batteries, very far from modern smartphones. But it is reasonable to expect that they will get better, and it is reasonable to expect that we will be able to emulate the brain with transistors and on a much bigger scale some days.
The second objection is that the demand for unskilled workers keeps shrinking. Our society has adapted to the industrial revolution by having a much larger skilled worker base, pretty much everyone can now read and write, college degrees have become the norm, etc. But there are brick walls to how skilled a society can be. A not insignificant share of the population just cannot become skilled. And even among skilled workers, only a fraction will always be capable to satisfy the demand for the "hotest" skills (engineering, software development, etc).
So I think that as that imbalance develops, we are bound to live with a structurally unequal society. And it doesn't have much to do with capitalism, it has more to do with the evolution of the technology outpacing the biological evolution of our species.
No, it's not reasonable to assume they will emulate brains.
Not even the experts are saying that.
They're not trying to emulate brains at all.
Also, even going back to von Neumann it has always been expected that the brain will be able to be built in hardware, simply because neurons exhibit binary behavior. They fire or not, and all the rest is massive interconnectedness and adaptive changes to their activation potential. There's lots of argument about whether reproducing the behavior of neurons will result in a 'mind' or be useful, but no one would argue that we simply won't be able to build a brain workalike.
Can you justify that statement?
I have literally seen a room full of people that lost their jobs because of software I wrote (admittedly, not alone, but still). The software was almost ready by that point and it was quite a disturbing experience that still bothers me up to this day to be honest.
It had nothing to do with AI and was quite a long time ago, but I still remember it very vividly.
What I have seen often too is bankers wealthy enough to be able to live on their savings for a long time, and therefore not willing to reconsider their career. And as a result staying unemployed for a long time before coming to the conclusion that they need to do something else.
I'm not saying this is fun and that I wouldn't mind going through it myself but this is capitalist's creative destruction at work. Not something you should feel particularly bad about.
Nobody has to be forced to reconsider his career (so his entire life..) when we just have to cut the link between work and salary that we learn at school.
The economy might adapt to change, but people aren't able to nearly as well.
You say "capitalist's creative destruction", but it's really "greed-driven devastation".
But I do think we (humanity) do ourselves a massive disservice when we push for automation for the sole purpose of reducing costs/boosting profits.
I also think that we need to remember it's real people being affected here. You say "spare them the unpleasant experience" - that's an almost comically insulting way to phrase it - it's really more like "spare them the emotionally traumatic and devastating experience of watching their future evaporate". You say "creative destruction", but you're really talking about lives being literally destroyed. And the key difference here is that (unlike previous time periods), there is no obvious future for these people who are losing their jobs to automation.
You rail against bankers not being willing to reconsider their careers, but what about cashiers at fast food restaurants? What about truck drivers? Train engineers? Miners? etc...
You say it's not something we should feel particularly bad about, but it absolutely is. I can't tell you how hard my eyes roll when I hear people talking in the almost kumbaya-esque way that I'm talking now, but we are all really in this together, and if we can't figure out a way to help people through the transition that is coming, well, what good are we as people?
Speaking so clinically about people's lives being destroyed does no one any favors. That's really all my point is.
If a company is able to automate its facilities to achieve 2X profits with half the workers, you'll have an higher overall GPD, but the created wealth will be shared by a reduced number of people.
People struggling to make a living won't give a damn whether TV has 3 channels or 300, they can travel cheaply around the world and there is a new smartphone model every year.
This has been sustainable up until technological advancement still allowed for enough decently paid unskilled jobs, but it's clear we are past this point.
We are running out of, as you say, "things for people to do".
I don't think these trends can or should be reversed (because they have undeniable benefits), but how to deal with the consequences is a real challenge.
Poor people in the United States, today, have a significantly higher standard of living than my grandparents, who were neither uneducated or poor for the time.
Poverty has moved way up Maslow's hierarchy.
This is a nice tidy test that is not entirely useful. You can't know anyone that didn't get hired into a job that didn't get created. Say a plant doubles output while destaffing by way of retirement. No one loses a job, automation eats more than 1 plants worth of jobs.
I'd be really interested to see sources for this since there is almost no accurate unemployment data pre-1840 and at least in the US it doesn't seem likely as what information we have tells us that during the period from 1800-1840 slaves made up over 26% of the workforce in the United States, peaking at almost 32% in 1810.
I'm assuming you're going by wage estimates from the UK over the period, but if you are those are really contested as a method for evaluating quality of life or actually real wages - many people at the time didn't earn inside the wage system.
I don't disagree that eventually cost of labor and general unemployment improved as a result of the industrial revolution - I just don't think it happened DURING the industrial revolution, at least not until its late stages.
Also, GDP per capita is a really bad way to estimate effects on the working class, particularly in a period famous for massive income inequality.
This system didn't work very well at the time (it bred slavery and warlordism, and produced a lot of waste, especially towards the end, when trade goods were abundant and population was running out); but it'll become increasingly more doable as capital becomes more productive. Magnanimity is a human instinct, and people already compete on status and prestige; just persuade holders of capital that supporting the unemployed/unemployable is a good thing, make sure they get personal feel-good-ness out of it, and persuade the unemployed/unemployable not to run up the red flag, and you've probably got a stable system.
Probably. I don't think we need to worry about slavery much, but warlordism could become a serious issue. Where potlatch-like culture exists, aggressive bids for the throne (or equivalent) tend to follow; the Earl of Warwick in the War of the Roses comes to mind. Still, even warlordism has potential to be less bad than Communism (more room for ambition, better management of capital) or Fascism (you can be a minority and not die) -- the Wars of the Roses had roughly zero impact on the common people of England. (See A Farewell to Alms for a detailed discussion of how England's wars, starting from a pretty early date, avoided messing up England's people.)
This is all hypothetical, though; I'm not saying that a potlatch orientation is the only possible solution, nor even that it's necessarily a working solution at all...
Taxes are at least predictable.
If you like getting giftcards at Xmas, nothing. But gifts from Santa or a potlatch are more fun, and more intimate. Choosing gifts forces us to think about each others' needs and desires, which is extremely pro-social.
I give self-replicating factories a higher probability than AGI. We'll probably have a 100% automated stack that can replicate and repair itself, as well as produce everything we need, all based on locally sourced and plentiful materials, before we have human level AI. It will be a manufacturing singularity. We should bootstrap automation in itself, like we bootstrap compilers. Object production would become similar to compilation, from source code to material object. We'd have optimized material compilers.
When we have a replicator, anyone can have one. Industry would become like agriculture, if you have the initial seeds, you can plant and then you bootstrap your crop.
If you think about self replication, on the whole, industry is a self replicator right now. But we need to make it fully automated, reduce its minimal footprint and make it less reliant on rare materials. There is a need for sustainable, smart materials and open source automation and AI in order to save the population from disenfranchisement. Capital and the means of production should belong to all of us, not just the 1%.
Fresh water, for example, is a scarce (and becoming scarcer) resource, and it's impossible to farm or live without it.
Agree with the comment about ownership. As technology has improved, more and more power has been going to whoever owns the capital.
The problem is that energy is expensive, and fresh water shortages will be worst in countries that are poor and can't afford these plants.
It's going to be some damn thing when everyone in the South takes your idea and they build up their civilization around free robotic labor, and then when we hit AI sentience and start ensuring the civil rights of robots the entire South has to be burned to the ground again to free the sentient robots from their owners.
We can still do the fun stuff of course, like writing and art, and get famous and all that.
Human labor has solved a lot of problems. The obvious shortcut to solving lots of problems is mimicking human labor. Given the opportunity to take a shortcut to solve a problem (especially at a profit) some manager in some business will do it.
And zero people can even describe how a sentient robot could be constructed, even with theoretical future hardware. "We'll have as many flops as some models say a brain has, and programmers are creative" is about as far as anyone can get, but I don't think that qualifies as a plan.
So, it's a fun thing to think about, but like all sci-fi it probably isn't the best basis for societal planning.
In order to perform at or above human level (in a general fashion), the robot will need a firm, comprehensive understanding of the world. Won't this unavoidably involve an understanding of the self as well?
I don't think any chips today have awareness. But it seems to me that you couldn't have general, adaptable intelligence without it.
But most modern AI is learnt from data, not programmed, how will we know if our AI in conscious?
We're not there yet and may never be, but since we do not know what consciousness is or what leads to it, we may stumble into it...
The rest you've said is a gross misrepresentation of the giveaway culture shared amongst many more tribes than just the Northwest ones. And of course, white academia will just refuse and say "our written documentation written by genocidal outsiders is way more accurate than what the cultures themselves say about themselves through their own oral traditions".
This must be an interesting book that somehow argues that wars that killed huge numbers of civilians somehow had zero impact. Looking at the criticism of the book it seems to also argue that there were no negative effects of colonialism.
Funny, I wonder if the millions of people killed in North America between 1500 and 1900 would agree with that.
Which, come to think of it, it's true.
Except some guilt maybe.
But just a couple corrections re potlatch society: slavery happened in the Pacific Northwest with and without potlatch; potlatch didn't cause slavery. The fishing grounds in this region were the richest in the world at the time (they collapsed almost a century ago under white rule), if you conquered a neighbor (and that was frequent) you could afford to keep 'em and feed 'em, at least for a long time until your population expanded. So slavery was practical in a way it isn't for nomads, say. Excess production was the norm. Unemployment was dealt with largely by exalting art, and devoting an immense amount of the GNP to that, with spectacular results over time. So most excess labor and production was actually devoted to art, not giveaways. I suspect that's the real lesson, although I'm a No Man's Sky devotee - it's a damn wonderful art generator, if you like landscapes, as I do. Maybe humans can't compete there, anymore, either! Warlordism in an age of drones will do little for unemployment; but it's also easier, particularly for democracies, if only your machines get bent up in a successful war of aggression. So it may happen more, but will not be a safety valve for unemployment.
That's a non-sequitur, you may as well say we're seeing this with the explosiong of freelancers and personal businesses.
Because that's really what those youtube personalities are, but that doesn't really address the issue of the older folks who have lost their jobs.
Warlordism sounds much worse than communism and fascism. Any culture that is sufficiently advanced to not be chaos and wars for the lords under walordism, would be much better without the warlords. Advanced societies with abundant resources and abundant capital do not need gatekeeping hoarders playing charity Olympics.
That's why we need to place the means of production in the hands of the people, not just of capitalists.
> What will it mean to work and contribute to society in the future of machines?
Self-supporting human society, a mix of small companies, skilled people and experts, using automation as well, that works to solve the daily necessities for itself. So it would have its own farms, fabs, schools and such, totally self supporting and self bootstrapped. People wouldn't need UBI if they had the means to directly make what they need. We're going that way - everything becoming cheaper, until it lifts itself by self replication.
Capitalism has actually provided far more avenues for laborers to own the means of production than any communist revolution has.
As automation gets better and cheaper, there will still be jobs for everyone - they'll just be performed by the robots they purchased for 1/10th, 1/20th or 1/50th of the median wage.
Lately I've really been grappling with how to accomplish this, and it is a Very Hard problem.
> Distributism has often been described in opposition to both socialism and capitalism, which distributists see as equally flawed and exploitive.
Entertainment will take longer, but I see no reason why it wouldn't be possible to algorithmically create. The lesser ones like Big Bang Theory is basically just a nerd reskin of every other sitcom Chuck Lorre has done. That could probably be churned out by a computer by the end of 2020. The more complicated entertainment will take longer to crack though.
Friendship is tough to define. So I'll give you that.
The first version could take the form of a pen pal over email.
and it's a good thing too, because if we are all out of work there'll be no good reason for us all to have so much intelligence so we'll me devolving back to the apes.
https://myprivate42.wordpress.com/2016/12/19/lets-shift-to-2...
as well as the obligation to consume, otherwise savings will be invalidated. Dystopia ahead.
First, history. History does not make our future look good. Humanity has gone through a few disruptive changes in the way society viewed work and societal contribution. The most recent was with the spread of factories and assembly-line production. The result of the adoption of those structures/technologies was not good. Society saw factory workers as 'not deserving' reasonable wages because 'the machines are doing all the work.' Factory owners were more than willing to pay their workers as little as they could get away with. And it turned out, that was very little. Entire families (including children) working 16 hour days 6 days a week and barely being able to feed themselves was commonplace. Society had to adopt many radical (in comparison to previous history and other societies, they don't seem radical now because we've kept them around long past any practical utility) changes to function like this. Everything from anti-sex attitudes, the creation of 'adolescent' as a distinct category of person, even changes in the architecture of lower class homes to give separate sleeping quarters for children and adults.
And that didn't get adjusted for until the New Deal in the 1930s. Back when someone couldn't simply say 'socialism' and completely shut down all discussion, the New Deal was actually possible. People today don't realize how 'insane' the New Deal really was. Think of it from the perspective of the factory owner. Previously, you would get an entire family, say 4 workers, 16 hours a day, 6 days a week, and you'd pay them just enough to eat and pay rent. After the New Deal, you were expected to pay 1 worker for 8 hours of work 5 days a week such a (comparatively) titanic amount of money that they could raise an entire family on it comfortably. Let that sink in. From 96 hours of work from each of 4 workers each week, 384 hours total, down to 40. That increases your labor cost by nearly an order of magnitude. Society re-configured their notion of what work was worth to address the idea that regardless of what tools are being used, the value the worker creates should play SOME role in compensation. And the world did not implode. But could anything proportionally similar ever be imagined to be acceptable to our modern society? I would argue absolutely not.
But there is good news. Maybe. See, this isn't our future. It's our present. Between 1950 and 1980, the median wage rose by 75%. Between 1980 and 2010, it rose by only 1%. Computers and automation technology caused average productivity to skyrocket since 1980. But wages were frozen. Not because of any economic reason, the median income of the top 10% rose by 475% during that 1980-2010 period. Profits grew and grew with greater acceleration, but all the benefit went to the owners, executives, etc. Society once again started feeling that workers did not deserve compensation where the value of the work being done played a role. The machine was doing the work. It just seemed too easy. You write some software and make the company $10 million, why should you get a "lottery winning payday"? Apple, Google, Microsoft, and similar employers that set the standards saw this coming and rigged wages in the computer market early on. This isn't a conspiracy theory, the FTC prosecuted them for it. They ended up settling for hundreds of millions of dollars in a huge class-action lawsuit a decade ago. But, really, you could argue it was inevitable. Society didn't see workers using computers as deserving of pay which increased alongside their productivity, so employers obliged society.
But something is different this time. Factories and assembly lines are expensive. And computers are cheap. And distribution, the whole reason we MADE factories and big companies in the first place, is effectively a solved problem. While companies have been getting rid of pensions, doing away with annual raises that exceed cost of living increases, cutting vacation time, expanding the work week, reflexively laying off workers to meet quarterly projections, etc... workers have become more than able to out-compete them. The drawbacks of centralization, which for a century were papered over with the invaluable ability to solve the problem of distribution, are now showing through more and more every day.
Employment is a two-way street. Employers pay workers because they create the value that is profit, and workers seek employment for things like a reliable income stream, the ability to settle down and make plans for the future, the ability to provide for their family. Except employers have welched on every single thing they ever offered to workers. They provide zero insulation from market fluctuations. They adopt practices that make it necessary for workers to jump jobs every 4-5 years just to keep up with the increases in cost of living. Etc. And it actually might have worked... if not for the fact that their ace in the hole, their one golden ticket, solving the problem of distribution (both of products and of work needing to be done, one to customers the other to workers) became worthless thanks to computers and the Internet.
Society changes slowly, and it certainly doesn't abandon century-long traditions overnight. But we're barrelling right for some proper Big changes. We're going to end up with the 'tools of production' (as Marx, who I've never read, would I think call it) in the hands of everyone, with a global market on our desk with commoditized distribution. Eventually enough people will be out of work as automation enables corporations to "reduce their primary cost center" and just trying offering their services online out of desperation that someone will build an excellent marketplace that actually works. And once it takes off, it will hopefully happen so fast that large companies don't even have a chance to react.
And I'm serious about that part. If they do have a chance to react... it could be uglier than you could possibly imagine. Wars have been fought over much, much less wealth. And things like the Internet are still primarily controlled at the last mile by large companies which will be directly and existentially threatened by this kind of thing. The government handed over a trillion dollars to make sure that the banking system didn't have to bear the consequences of their actions, what might they do if 50% of the Fortune 500 were facing insolvency?
So of course there will be a global pressure on salaries. It is just market economics. (You can't coordinate a "New Deal" in countries all over the world at the same time.)
My point about that is -- it isn't all bad. Literally billions of people have left the utmost poverty. It just sucks for the ones of the old "middle privileged" level, like us.
If it was just that, I would cautiously be positive, but I agree -- this new level of automation implies "Big changes", which are scary. Most revolutions seems to end up with a large part of the population dead or in slavery.
Communist parties were gaining steam back then, and we were under threat of having our own Red October if something extreme wasn't done to blunt the effects of the Great Depression.
The truth is he didn't bother to try and sell it to the people in power because he didn't have to. It was readily apparent that if something wasn't done America might not make it. At the time unemployment rate was at almost 30%. Most american's didn't have access to their bank accounts as many were insolvent. Literally the country was on the brink. Congress was going to do whatever it took to drag us out of that hole even if it meant completely rewriting the rules of well everything.
I don't know that any of that is possible today. The regulations we have in place largely prevent us from being able to get back to that sort of dark economic time. And in my opinion it will take either that sort of global catastrophe, or one of our current major parties to fall apart to enable the sort of cooperation in congress New Deal level policy requires.
The real issues won't change though. Those people who are considered rich will be rich because they own land in nice areas, or have a military-backed monopoly on natural resources.
The first issue is due to population size, and the second issue is an eternal issue of war that can't be avoided.
Best way to accomplish this, I think, is to have a Basic Income. Give everybody enough to live, and increase that the more machines take over our work.
In our current economic system, we get the former. We need to change the system so it turns into the latter.
We are suffering today not because of technological progress, but rather the contrary: the rapid technological progress of the 20th century that brought tremendous economic prosperity to humanity has finally come to a grinding halt. Let's stop denying this. The stream of lifechanging breakthrough inventions of the 20th century, from A (antibiotics) to Z (zippers), have ended. As a result, we now suffer from secular stagnation. [1]
This is why people can't get jobs. The answer is Keynesian spending, and the most fair kind of Keynesian spending is basic income. Call it a "Keynesian dividend".
[1] http://larrysummers.com/2016/02/17/the-age-of-secular-stagna...
IMHO these articles are fake news, journalists love beating the Automation is killing jobs drum.
IT certainly is changing and outsourcing is swings and roundabouts. But there's more work than ever.
Suddenly the American oligarchy loses its faith in automation once you discuss potentially inflationary increases to public spending.
If we really are in an automation renaissance public spending could be ramped up massively and the potentially inflationary impact would be offset by all of the cost savings from replacing people with robots.
If.
Oh god, stop with this in every article.
This isn't "fake"; it isn't even "news".
News is either fake or not. A researched opinion piece isn't "fake" because you disagree with the conclusions. It might be wrong, but it isn't "fake".
A worldwide monetary policy change that keeps real interest rates sufficiently negative to be in line with private market rates is the only real solution.
I agree, but this only leads to more questions: why? And is it permanent?
I think we are in a temporary lull. Finance people, in the vain of JP Morgan, have gotten very good at controlling the social processes around both physical resources and human output. They keep most people tired and scared to change jobs, so they can be made to focus on production. The rest who are actively problem solving despite those pressures, the finance folks try to steer them towards a maze where big bets are made and equity changes hands often.
We were bison, and now we're cattle. There's more meat, but less frolicking, and no semblance of any kind of ecosystem that produced so many bison-y moments.
Is it forever? I don't think so. We're only fairly recently getting lots of brains trying to discover tools for doing things outside of finance. And still, it's a trickle. Look at all the beautiful brains on this message board, whose output is getting fed into the feedlots of VCs and Paul Graham.
Our rapid technological progress in the last 2 centuries was an anomaly. We have 200,000 years of history as a species, most of which did not have nearly the same level of progress. The burden is on to you to explain what made those 2 centuries so great.
Automation today is now destroying service economy jobs, but it's not clear, at least to me, what new sector is ramping up to pick up the displacement slack. The tech sector isn't going to have enough jobs for everyone getting automated away, and this time around automation is unique in that it's attacking pretty much everything all at once. Human drivers are going to be displaced by autonomous vehicles at the same time as financial advisers are being displaced by wealth allocation algorithms at the same time as warehouse workers are being displaced by picker robots, etc.
Patching all but the simplest software bugs will require real AGI. Understanding a bug and the side effects of a patch goes way beyond current AI technology, and there is still no defined R&D path which is likely to produce an AGI.
The narrative will follow the "drug war." When it's the poor people (factory workers::blacks), they're just "losers" who need to try harder or should be locked away for decades. When it's happening in upper middle and middle class suburbia (opioid epidemic::white collar automation), then it's a "public health crisis that commands our empathy!"
How many programmers would it take to write Windows in Assembly?
That's exactly right. We've been creating work to keep people busy for 200 years in many more ways than you mention. Consumerism. Marketing. Bureaucracy. We have one industry that makes people sick (food) and another industry that makes people healthy (healthcare). The twin peaks of the sickness and health industries can grow indefinitely, especially in the USA. We have industries that make plastic casings to wrap goods, just so we can sell goods to consumers, who unwrap and immediately dispose the plastic, so that another industry (sanitation) can take them away.
We basically have a society which continuously breaks glass windows and then fixes them. It's all, in some sense, pure dead weight, just to keep people employed.
So why should this stop now?
Think of the additional economic boost from treating these people in hospitals, or even better, prisons! Countless "jobs" will be created!
Health and food are actually a great example. We've already automated away most farm jobs (in the US at least) and a lot of the food production & prep jobs for junk food type things. And now we're starting to see in the healthcare space that software can reduce loads on doctors, for example ML-based interpretation of X-rays, automated patient record tracking, etc. Fitbit is the tip of iceberg for what's possible in at-home health monitoring based on software instead of humans. I would argue that while the sickness and health industries can grow indefinitely, the number of jobs in each cannot.
Fine, I can buy that. But stagnant productivity alone doesn't generate low employment. How do you explain low employment then?
We have already done things like basic income in the USA:
- Economic Stimulus Act of 2008
- Earned Income Tax Credit
Please provide some sort of argumentation or data.
Nah. The mother of all tech progressions, AI, is just kicking in.
There are other considerations of course, such as proximity to resources, transport hubs, and markets.
Ultimately though, this doesn't change the fact that automation is mostly beneficial to capital owners and not the wider workforce, so capitalism's tendency toward inequality (as described by Piketty) still applies.
But this doesn't really bring jobs back, since they'll just be automated.
That's not to say there aren't other benefits aside from job creation, though.
There are also a lot of engineering jobs designing the hardware and software of these automated factories.
Seems like there typically are fewer jobs and the total compensation for labor is reduced when manufacturing is automated (why else would the company pursue such a course of action?)
Automation will continue, but labor will probably never be completely replaced within our lifetimes. I think it makes sense to have factories and make physical goods in the US. It is easier to bring hardware to market when the production is in the same timezone. I think it leads to a more robust economy even if there is heavy automation. There are still ancillary jobs associated with factories.
If we can attract new factories then it could be a net win, but someone will lose out (not china, but Bangladesh, India, Vietnam, and other countries that are trying to break into middle income).
Since we already produce about everything in the amount we need (and sometimes much more than we need), there is no point in increasing production (it would not be sold, consumed), so productivity gains will translate into cost reduction, i.e. less jobs.
Yeah, that's right. There would certainly be other benefits (and a limited number of jobs created) that are more indirect.
The real issue around automation in my opinion is how we as a society manage it. If we stay the course, it is clear that capital owners stand to benefit most (almost entirely I would say). We are already seeing problems with rising inequality due to diverging returns from capital and labour. This is going to exacerbate that hugely.
I'm generally in favour of capitalism but I think we are going to need to revisit 'the social contract' that allows capitalism to exist in exchange for a share of the rewards being used to ensure that society's basic needs are met. Just as when it was first negotiated by Bismark, this is going to have to be an initiative of the capital owners if serious unrest is to be avoided.
I'm hopeful and think that it is possible. Rich people are not inherently sociopathic. A good example of this sort of compromise in more modern times was the reaction of Kennedy snr when the US faced economic and social unrest. Partly through correspondence with Keynes he was convinced that if himself and his peers didn't take a hit for the good of wider society then they stood a good chance of losing a great deal more.
Every so often the social contract needs a reset (e.g. Bismark, the labour movement, the new deal, etc). Automation is going to make this more urgent. The one thing I am sure of is that it will be a political solution and not simply an economic one (e.g. less regulation, more free markets, etc).
I think that inequality will increase but not because of owning capital. Rather because this automation will shift demand away from unskilled workers and into software developers, marketing experts, etc. And I do not think that the supply of skilled worker can adapt fast enough, if any.
The good news (for the US) is that China will be the main casualty. In a way the US has already largely paid this tax to automation by outsourcing factory jobs to China. It is chinese factory workers that are going to be automated the first. Some low skilled jobs that couldn't be outsourced offshore like construction, taxis, etc will suffer but I'd argue a large part of the impact is behind us.
Obviously the tax paid to outsourcing is also having Trump at the white house.
If it is not capital ownership, then some other inequality mechanism is at work if the automation trend is true, considering that as automation scoops the obvious and not-so-obvious patterns of workflows in the economy out of the labor market, proportionally none of the benefits accrue to those who are left creating that automation. Think about that for a second.
If you are one of the few who create the automation, then by definition what you are doing is currently not amenable to automation. You are creating a fabulous wealth machine with the automation, but by and large, unless you own a capital stake in that exercise, you will not see compensation proportional to the value you bring to the table. It's a curious dynamic to me, because due to classical economics, I would have expected capital to flow more readily to those whose work could not (yet) be automated, to accelerate delivery towards the proven benefits of automation.
It's easy to assume and take for granted that the US will create all the automation technology. However, I think it's more probable that China will move up the value chain and be the ones to own and develop these technologies, especially in manufacturing as China already has a super strong manufacturing ecosystem and talent (think Silicon Valley and Software).
If China owns the technology, the wealth would be in China so less will be available to spread around for basic income in developed countries like the US.
"No candidate talked much about automation on the campaign trail. Technology is not as convenient a villain as China or Mexico, there is no clear way to stop it, and many of the technology companies are in the United States and benefit the country in many ways."
A lot of US jobs have been lost to globalization, but in the long-term that won't compare to the jobs lost to automation. As a software developer this seems painfully obvious to me, and I would hope most politicians are intelligent enough to see it as well. Unfortunately none of them have even talked about it, much less proposed solutions to help displaced workers.
Technology is a MUCH more convenient villain. You can roll back 1% friendly trade agreements with China and Mexico. You can point pitchforks at the 1%. You can't roll back technological advancement and it's pointless to try.
You achieve nothing by being against technology except looking like an idiot, which is precisely why articles like this are trying to convince everybody that technology is the "villain".
Because god forbid Lloyd Blankfein and his cronies start being seen as villains.
>A lot of US jobs have been lost to globalization, but in the long-term that won't compare to the jobs lost to automation. As a software developer this seems painfully obvious to me,
Funny. As a software developer it seems painfully obvious to me just how far off we are from this supposed goal. All around me I see developers who shirk at even automating their own tests while getting excited over vaporware.
If you want manufacturing jobs to stay in the US, regulating automation would be far more effective than trying to prevent companies from outsourcing jobs. (Incidentally, a lot of jobs that have been outsourced probably would have been automated if outsourcing had not been possible, and that's what opponents of outsourcing are missing.)
I'm very much against a policy to restrict automation, but it's worth pointing out.
Once that works, California loses jobs about 2.1 million illegal immigrants.
Automation and industrialization over the last 150 years already meant we went from 120h weeks on farms to 40h weeks in offices and factories, while living standards increased. Measured in 40h/week jobs per person that means we lost 2 of 3 jobs in the industrialization! The world didn't end. And now everyone is worried that if we go from 40 to 30h or 20h weeks that will be a disaster?
We can't afford to have a large fraction of the people unemployed and the rest still working 40h weeks in ever fewer jobs of course, but that's a problem societies will hopefully solve.
While we still measure a person's worth out of how much they make, this is going to pose an enormous problem for all of us. We can't just take away people's livelihoods for the benefit of a few (us, to be honest).
There will be a violent reaction, as there has been every time this situation has arisen throughout history.
But how is shifting from e.g. production to services taking away peoples livielihood any more than shifting from farms to industry was?
> While we still measure a person's worth out of how much they make, this is going to pose an enormous problem for all of us.
Aren't most developed countries already either speculating in, or already effectively using shorter working hours now than say just one or two decades ago? Working less seems like the natural way to go. https://en.wikipedia.org/wiki/Working_time#Gradual_decrease_...
Edit: I should add: I'm optimistic about this in societies with a high degree of labour organization, a flexible/agile political system, reasonably strong welfare states and a population that is positive towards working less in general.
Simply put: if you are in an OECD country and using kilograms you'll be fine.
And service jobs are not immune to automation either, eg Eatsa.
As working hours drop and wages tend toward minimum, we do indeed end up in a bad situation.
There is also a (imho) great SF short story by Marshall Brain about this: http://marshallbrain.com/manna1.htm
Another big thing that drives automation is the stock market. The stock market provides a bunch of money to companies that can show earnings growth. Used to be you could just expand globally. Well, all the big brands have done that now. So the next step is to cut costs. How?
Automate.
Without the stock market this might also happen but at a slower pace.
[1] https://www.youtube.com/watch?v=7Pq-S557XQU
Anyone who thinks that automation isn't a big deal want to respond?
I'm also curious how we assume with such confidence that everything will be fine since we survived the Industrial Revolution when there was an order of magnitude less people on the planet in the 19th century?
In the meantime, let's bring the manufacturing chain back here so we can build up the automation expertise ourselves, and, while we wait for the End of Work, stop making the middle of our IQ bell curve compete with countries that don't give a shit about their environment, the concept of intellectual property or the safety of their workers.
We can figure the automation problem out later.
How? There's obvious long term financial value in having local supply chains and manufacturing expertise. Therefore state capitalist systems will subsidize local manufacturing in the short term, while anarcho-capitalist systems will flock to set up shop there to capture convert those short-term subsidies into short-term profit.
Do you have a solution that will work to build high-tech manufacturing infrastructure in the United States with no corporate incentives and no US government intervention? (I mean besides tweeting brags about job creation)
The US corporate income tax rate has been a bad joke for a long time. Countries like Sweden at 22%, understood decades ago that it was bad economic policy to have a high rate. You tax the income of the wealthy as the offset. The US for example already has an extremely progressive personal income tax rate. For several decades the corporate income tax rate has been falling around the world; the world's average rate has fallen from near 30% to 22% in the last 12 years. The US isn't competitive, you see that in how we've been bleeding pharma & biotech companies off to Ireland. Over time, you start seeing R&D & operations offshoring because of that. Give it time and you've artificially created a lot more global competition through bad policy.
If you're a small to mid size manufacturer in the US, paying a 30% effective income tax rate is brutal if you want to compete globally while everyone else is paying far lower rates. Germany for example lowered their top rate by about 9 points a decade ago. Finland and Iceland have a 20% rate. Korea is at 24% and China at 25%. The European average is about 20%.
If you're talking about cutting the statutory rate while also eliminating deductions and simplifying the code, I'm all for that (ditto for personal income taxes). A huge complicated tax code benefits big corporations more than your small business owner, because tax lawyers and accountants are largely a fixed cost.
But let's not keep repeating the misguided fact that US corporations pay unusually high taxes -- they simply don't.
[1] http://www.forbes.com/sites/taxanalysts/2015/03/25/the-truth...
The observation that a complicated makes the big guy part a lower effective rate than the small guy should be much more popular, I wish someone established a good name for it.
That's a great way to destroy the environment and increase corporate profits, but I was asking about how to get companies to ignore the profit of offshoring; how to convince them to act in direct opposition to their shareholder's interests, and not take advantage of the subsidies provided by state capitalist systems and locate manufacturing overseas.
All cutting the tax rate is going to to is make offshoring more profitable. At no time in US history has a tax cut been correlated with business growth and job growth. Quite the contrary, The Bush tax repatriation holiday gave many businesses the working capital to invest in closing plants and moving factories overseas, resulting in the net job losses in many industries. But if you have hard data on examples of cutting taxes below a top marginal rate of 40% in western countries having any effect other than increasing income inequality, I'd love to see your peer reviewed study.
And is there a particular regulation that you can think of cutting that would have more of lure for for domestic manufacturing than China dumping trillions of dollars into currency manipulation and building infrastructure and manufacturing subsidies?
FYI: The US produces way more CO2 per capita than China - about double, IIRC.
edit: source -http://data.worldbank.org/indicator/EN.ATM.CO2E.PC?locations...
Mass automation is undermining our democracy in a very specific way: it's acting as the ultimate "resource curse." https://en.wikipedia.org/wiki/Resource_curse
"Countries with an abundance of natural resources, specifically non-renewable resources like minerals and fuels, tend to have less economic growth, less democracy, and worse development outcomes than countries with fewer natural resources."
Scholars debate the causes of the resource curse, but one popular theory has to do with the way autocrats fund themselves relative to democracies.
Autocrats, it turns out, need a lot of wealth to pay their cronies. No dictator rules alone; they need someone to run the military, someone to collect the taxes, and someone to enforce the laws. Those people have to be paid, and handsomely, or they'll overthrow the dictator (or just allow the dictator to be overthrown). This is called "selectorate theory" and this video is a great introduction. https://www.youtube.com/watch?v=rStL7niR7gs
Oil wealth, specifically, undermines democracy because when autocrats have access to oil wealth, they don't need to depend on their citizens very much. (Indeed, many oil-rich autocratic countries just allow other countries to come in and drill it, keeping local labor entirely out of the loop.)
Resource-cursed autocracies tend to democratize when the oil wealth runs out and they need to rely on the people's productivity to deliver wealth to cronies. When autocrats are forced to allow people to educate themselves and communicate with one another, democracy ensues.
It can work the other way, too. In every democracy, there's a group of folks asking themselves a question: is now the time to try a coup, to replace democracy with an autocracy? As the value of capital increases and the value of human labor decreases, the advantages of staging a coup become more and more enticing.
For years we've thought of human labor as the "ultimate resource." But it turns out that human labor isn't the ultimate resource. Robot labor that's just as good if not better than human labor is a resource beyond any we've ever seen.
But that means that we're discovering/inventing the ultimate resource curse.
We might use automation to fund universal basic income, or a class of elites could use it to undermine "unnecessary" citizens (the "unnecessariat"), establishing a corporate fascism.
When the government depends on human productivity for our tax base, the government needs to keep us all well-educated and healthy. But soon, government won't depend on human labor.
"Is now the time?" they're asking. And, increasingly, the answer is "yes."
The biggest worry here is the automation of law enforcement. Once we're policed by robots, there's nobody to rise up against the dictators.
Maybe we should ensure that every person has one robot to do their work for them, and that's it. No robots for governments and corporations.
This. This exact thought occurred to me about a week ago -- to my shame, a full year after I read The Dictator's Handbook.
The natural concentration of robot ownership in the hands of a few creates the same autocracy dynamic as described in the book for oil and other natural resources. Also, robotic factories can be seized and controlled by an autocracy, just like oil wells. What's really scary about this is that threatens stable democracies, like the USA.
These are two issues that will actually help with both creating jobs and ensuring decent pay. Nobody really knows what will happen in the future. If there are robots that can do all of our work, who says that only extremely rich people will be able to afford them? Which technologies have stayed permanently unaffordable where there is consumer demand? Not that many.
We approach 2017, we know climate change is a real threat, we know automation is going to be a big problem, we know we're consuming the world's resources at an alarming rate and there's 7.5 billion people on this rock...
I get the feeling the future won't be this Utopian scenario that some people are dreaming of.
We've done that. Starting with the first language after Assembly.
We'll have to solve that paradox and change our mindset on life. Philosophy time.
The revolt in late XIXth century in Europe were based on the generalisation of the steam engine that were favoring without any other merits than birth the wealthy and ripping the craftsmen of their jobs.
Diesel made an engine so craftsman could compete vs steam engine. The german government pre-empted his invention so that it would not disrupt the dominant position of big corporations...
Automation is not the problem, it is unfair access to automation decided by capital.
Allow me to explain. Here's a youtube movie of some robots in a Mercedes factory
https://www.youtube.com/watch?v=VreG1iC65Lc
There are thousands of movies like that. Lots of robots doing a lot of work, and apparently displacing lots of jobs. But there are some jobs in the background. Someone has programmed those robots. But aren't those programmers (super-)highly skilled?
I don't know for sure, but it feels to me a robot is programed like an Excel macro: you record the moves, and then edit what you recorded. For an Excel macro, both parts are done (generally) by the same person, but for industrial robots, they are probably done by different people. The person who is being recorded needs only know how to manipulate a mechanical arm with a controller, that is probably not very different from a PS4 controller. The person who edits the macros, creates procedures, then combines them, optimizes them, creates tests, etc, etc, that person is a highly skilled individual. The other one not so much. The difference here is the capacity for abstraction.
Now, one could say that both these jobs happen only once, and then the robot performs the jobs thousands of times, so you have 2 new jobs (one high skill, one low skill) displacing thousands others. Alas, the programming, maintenance, re-programming, upgrading, not to mention construction of the robots still needs a small army of people. But overall, I agree that it's possible that introducing these robots resulted in a net job destruction for Mercedes and its suppliers (but not entirely sure; Mercedes was most likely more concerned with the quality of their cars rather than their cost of production).
However, not every manufacturer is similar to Mercedes. A lot of manufacturers produce things in smaller batches for example. And now you start to look at the trade-off of putting the automation in place:
If a batch is small enough, the initial investment in automation is not recouped. If the batch is huge, the automation is a no-brainer. But there's a huge middle ground where only some partial automation makes sense. And part of that automation that makes sense is to do things using some mechanical arms moved by some PS4-type controller.
I think in the near future, a lot of the "mechanical" type of work (welding, cutting, soldering, hammering) will become remote-controlled. It might look like high-skilled now, but the job of a machinist was considered high-skilled 50 years ago, and essentially they are the same.
Now, the farther you are from the moving or hot or electricity-conducting parts, the farther you are from danger, and the more people can try something as a hobby. More importantly, the barrier of entry for a given (low-skilled) trade goes down, so people will be able to retrain themselves more quickly. And these people will be able to create more quickly.
Just like blogging has created millions of (unofficial, but money making) jobs, the advent of remote-controlled machines will create millions of new (low-skilled) jobs. And these remote-controlled machine will show up exactly because of the need for automation.
If automation doesn't result in absolutely massive savings on employment, it just won't happen. You're army of people is more like a small platoon -- the army being entirely displaced.
But I think there is something to be said about automation allowing for creation to become more accessible. It might make entire categories of small-scale products designed by small teams possible in ways that were not possible before.
Good thing is though, machines dont need CEOs - so the last rat on the ship, might have the flagpoles cotton as towel.
That is unless they can invent so many boolshit jobs, that the hamster wheel keeps on churning.
He means it. He brought in the director of the Tapei Zoo for advice on how to manage animals with different temperaments.
We're going to have to find ways to help people stay busy/occupied. Otherwise, society will just turn nasty.