tl;dr Employing robots and AIs overall to do almost *anything* may soon be cheaper than employing most humans, similar to how we junk old computers because they use too much power per MIP even when they still "work". Also, the demand for market-supplied goods and services is limited as people move up Maslow's hierarchy of needs and want to do more things for themselves and want to do creative things of a (generally) less materially-intensive nature. Thus mainstream capitalism is about to face divide-by-zero errors.
=== Two key assumptions of mainstream capitalist economics
Your insightful comment really gets at the heart of the controversy (from a mainstream economics viewpoint) and so deserves a significant response. I agree that we can see many of the material aspects of US society today as a "victory for automation". Still, as your post implies, mainstream economics includes two assumptions, and if either of these assumptions becomes false, mainstream economics will falter (although it may press on anyway at the cost of great human suffering).
These two assumptions are:
1. The assumption that most human labor will always have significant exchange value in the market regardless of advances in AI, robotics, and other automation (as well as other trends like voluntary social networks, accumulation of infrastructure and know-how, cheap energy like from fusion power, and so on). Significant means at least enough value to purchase at least the bare minimum to survive (given almost all the land has been privatized or otherwise enclosed preventing subsistence through hunting, gathering, and gardening). Further there is the assumption nowadays that the exchange value of labor in the USA by most people willing and able to work is high enough for more than subsistence to achieve at least a "middle class" US life or better.
2. The assumption that demand for goods and services is infinite for any given population (or at least demand grows faster than supply over the long term), thus ensuring there is always a need for more labor no matter how much automation can meet previous demands. There is also the assumption this excess demand for goods and services beyond supply will be great enough to prevent a race to the bottom in wages through competition regarding supply and demand of human labor. In short, without major government intervention like a basic income, a capitalist system as we know it needs to continually expand to be viable for human survival if most humans require employment to survive.
Let's see if either of those assumptions can be debunked. It only takes debunking one to show how mainstream economics is failing in the 21st century. Debunking either would mean there would be zero wages for most humans or zero new net demand, and some mainstream economic equations will blow up with divide-by-zero errors in such a case.
== Assumption 1: Most human labor will always have market-place value
Debunking the first assumption about human labor always having significant exchange value is what the video is mostly about. And it is enough to invalidate mainstream economics as we know it in the USA. A similar point was made in "The Triple Revolution Memorandum" from 1964, although for various reasons I mention on my site was ahead of its time in its predictions. As an analogy, many Slashdotters have older computer around they don't use anymore. Those older computers are generally turned off because compared to later computers, they are slower, are noisier, consume more power per MIP, can't be updated for newer security threats, have limited memory, may have more complex UIs or maintenance procedures, and so on. It is easier in most cases to just run VirtualBox on a multi-core machine with a task-specific VM or to use small cheap recent dedicated hardware for functions like routing or firewalls or file serving. Or Slashdotters might just do more with cloud computing that others maintain. The CGP Grey video is about the trend towards most humans becoming like old computers as far as employability in a modern economy relative to special-purpose robotics, AIs, and other automation as such get better at mental labor and visual pattern recognition. Yes, perhaps old computers could do the job slowly, but managing old computers could be more trouble than it is worth given the costs and risks and given that the quality of the results may be reduced.
How does that analogy play out for real human workers?
Our current economic system produces a huge amount of goods and services but has trouble distributing them in a healthy way. That is because for most people the right to consume is linked with the exchange value of their labor, which varies widely. An alternative would be to see at least part of the right to consume as being a human right from being a citizen (like talked about by C. H. Douglas and Social Credit or President Franklin D. Roosevelt and "Freedom from Want"). Instead, in the USA, beyond some special cases related to disability of some sort (youth, age, illness), we as a society generally only grant the right to consume without exchanging labor to those few who for some reason can claim financial capital ownership. However, the history behind so much (not all) capital is that it was arbitrary granted centuries ago by some long-dead King as land grants or monopoly patents or similar and passed down through the generations (who, may have done some work deciding on who to trust to employ to decide how to invest it). Capital then grew like a snowball, with the rich getting richer, since it generally takes money to make money. Even with many exceptions (Steve Jobs?), the fairness of the distribution of capital can be questioned. Or, at least, capital could be taxed on that justification -- although generally most of capital isn't taxed.
To obtain goods and services other than via exchange labor such as via government welfare payments (from redistributive taxes, royalties, or inflationary printing money) is looked down on which produces stressful low social status (unless your are "retired" or "disabled" or a "student" -- or "wealthy").
The exchange value of labor for most people in the industrialized world is diminishing relative to the value of capital and the value of some few highly-valued skills (like those of automation engineers). This is shown by flat real wages in the USA for thirty years. Yes, health insurance payments have risen as part of total compensation, but not to much visible benefit for most workers. Other factors diminishing the value of paid labor include volunteer social networks (as have been displacing newspapers and such via web forums like Slashdot) and also potentially cheaper energy (which can brute force some things like recycling otherwise requiring human finesse) and greater collective know-how (which can make products last longer or work better and so reduce demand, like modern rust-proofed cars that last longer with less maintenance). Thus real wages will likely continue to stagnate or even decline for most people because there is no great demand for their labor given these trends, and labor competition ultimately produces a race to the bottom for wages as for any market commodity.
Without minimum wage laws and other public assistance, we are perhaps already at the point where a significant percentage of the US population would not be able to feed themselves from jobs because their labor has little economic value in our competitive economy which included many desperate out-of-work people willing to work for low wages. However, raising the minimum wage just provides more incentives to automate human jobs out of existence faster ultimately as an economic death spiral for most humans (absent strong government intervention such as in much of Western Europe).
Capital-heavy automation tends to concentrate wealth like Marshall Brain talks about. Despite claims about the wealthy as "job creators", in practice, absent proven consumer demand for new goods and services, the wealthy tend to hoard their wealth or play with their cash in a zero-sum casino economy of the "FIRE" sector (like discussed in "Money as Debt II"). This video by Richard Wolff, even just in the preview, talks in depth about flat wages and increasing productivity for the last few decades (although I don't agree with all the proposed remedies there):
As you say, US America is good at producing individual material wealth. So, with many more goods and services to buy and flat wages, there is a cash crisis in the real economy for most people (despite vast amounts of dollars circulating). This caused an economic recession for real goods production and middle class services due to lack of apparent demand from lack of cash (since the market only hears the needs of those with money).
The important point though for this discussion is that if US most workers' labor had really been in demand in a big way, their wages would have risen instead of corporate profits having risen. That is a fact of capitalism. That's the same way we know H1Bs are a scam -- because there has not been a major rise in wages for US tech professionals over the last decade or two as one might expect if there was a major shortage of tech labor. We are then left debating whether weak demand for most US American labor in a growing economy is due to automation, globalization, political trends, or all of the above. Even if automation was not the biggest factor (debatable), the issue is also what will happen going forward?
As radical as the conclusion may seem about the unemployability of most humans, I am merely suggesting that the trends we have been seeing in the US economy for the past three decades are going to continue and spread globally. Mainstream economists, if they admit these trends at all, are IMHO more radically suggesting these multi-decade trends will suddenly reverse themselves by some mysterious natural functioning of markets. To me, that is like the denial-istic discussion between the two horses in the video. Still, as high priests of capitalism, most currently-employed academic mainstream economists will probably have job security for a long time regardless of the reality of what they profess.
Big name Nobel-prize-winning economists like Paul Krugman sometimes flirt with accepting reality, but then back away from championing the implications. For example, from last year:
"I've noted before that the nature of rising inequality in America changed around 2000. Until then, it was all about worker versus worker; the distribution of income between labor and capital -- between wages and profits, if you like -- had been stable for decades. Since then, however, labor's share of the pie has fallen sharply. As it turns out, this is not a uniquely American phenomenon. A new report from the International Labor Organization points out that the same thing has been happening in many other countries, which is what you'd expect to see if global technological trends were turning against workers.
And some of those turns may well be sudden. The McKinsey Global Institute recently released a report on a dozen major new technologies that it considers likely to be "disruptive," upsetting existing market and social arrangements. Even a quick scan of the report's list suggests that some of the victims of disruption will be workers who are currently considered highly skilled, and who invested a lot of time and money in acquiring those skills. For example, the report suggests that we're going to be seeing a lot of "automation of knowledge work," with software doing things that used to require college graduates. Advanced robotics could further diminish employment in manufacturing, but it could also replace some medical professionals.
So what is the answer? If the picture I've drawn is at all right, the only way we could have anything resembling a middle-class society -- a society in which ordinary citizens have a reasonable assurance of maintaining a decent life as long as they work hard and play by the rules -- would be by having a strong social safety net, one that guarantees not just health care but a minimum income, too. And with an ever-rising share of income going to capital rather than labor, that safety net would have to be paid for to an important extent via taxes on profits and/or investment income. I can already hear conservatives shouting about the evils of "redistribution." But what, exactly, would they propose instead?"
Government policy that gives newly created money first to the FIRE sector (banks) does not solve the underlying crisis which continues to unfold in the USA. Real jobs might have been created if the cash had gone to the average US American, who is a real "job creator" via demand, but that is not what happened. As in the video "Capitalism Hits the Fan", that cash crisis had only been kept at bay for so long in the USA by the wealthy loaning cash to workers (rather than paying them). But that approach failed around 2008 as workers overall reached past even the most optimistic (sub-prime) borrowing limits.
We are likely about to see the same thing repeated with student loans, BTW. That is because of the lack of good first jobs that can ultimately repay the opportunity cost of several years out of the work force in college ($100K+?, assuming there were jobs) and then $50K+ or so of loans on top of that for many college kids. However, student loan cash has until now helped prop up spending by young people and kept them out of the workforce. If more young people question the value of college and grad school as an "investment" based on those trends, and instead try to go directly into the workforce and learn on the job (like 18 year olds used to in the 1950s), we would see higher unemployment numbers.
Automation (including computer-supported bureaucracy) supports wealth concentration (like with Walmart or Google or Facebook or Bridgewater and so on). Concentrated wealth in the USA can remake employment laws to be even further against the average worker (short of a big political movement). Concentrated wealth also means the average worker is in an even worse position to bargain for wages. Marshall Brain explains that with diagrams in "Robotic Freedom" here:
At some point, the market-based economic value of most human labor compared to automation may reach the point where it makes no sense for companies to employ humans even if the human labor is free. It might not even make sense to hire humans even if money is paid to the company by the government to hire someone (for "training"). This is because every human hired requires minimum OSHA standards of heat and light and ventilation, which can be more expensive than lights-out factories. That's been a selling point of Kiva robot systems for warehouses -- that they reduce the amount of space that needs to be air-conditioned and lighted for humans. Also, humans employees pose risks of lawsuits (accidents, harassment, discrimination, etc.). Humans are also unreliable (sickness, addiction, psychology, tiredness, boredom) and so may make mistakes causing expensive scrap. Would a big hedge fund like Bridgewater employ humans to draw financial graphs by hand instead of doing that via software plotting data from a database, even if Bridgewater got $100 an hour from the government to employ that human? It's just not worth it to Bridgewater, given limited management attention to supervise people and a need for speed and accuracy. Even FoxConn is bringing in robots because they are cheaper and more reliable than (cheap) Chinese workers. While it is true some workers may make suggestion about how to improve processes (Japanese companies excelled at making the most of that), vast networks of connected AIs may eventually be able to do the same -- and in any case, that's what a small group of thoughtful engineers can do.
== Assumption 2: Overall demand is unlimited in the market place
The second assumption about unlimited demand is not substantially addressed by the video. It is harder to debate, and you make good points about it. Still, I've questioned it on my own pdfernhout.net website though. Essentially, there is a law of diminishing returns for more goods and services for any *healthy* human. Granted, the needs of unhealthy humans may be different, like illustrated in one of the stories in Frederik Pohl's "Midas World" collection:
""The Man Who Ate the World" (originally published in Galaxy in 1956) tells the story of Anderson Trumie who had a scarring experience in his childhood, before Morey Fry changed the world [with cheap fusion energy]. All he wants is a teddy bear, but his parents' lifestyle of frantic consumption won't allow him to have one anymore. As an adult, he is a compulsive consumer. He has taken over North Guardian Island and he is putting a burden on the local infrastructure. A psychist, Roger Garrick, with the help of Kathryn Pender, find a way to heal Anderson and end his exorbitant consumption."
Like the Raffie song:
"All I really need is a song in my heart
Food in my belly and love in my family
All I really need is a song in my heart
And love in my family
And I need some clean water for drinking
And I need some clean air for breathing
So that I can grow up strong
And take my place where I belong ..."
The limits on human needs (or material demands) is in part due to Maslow's hierarchy of needs (which you indirectly mention) and the need for a certain amount of variety in a healthy life (including a need for non-material things like community and a need to do things and make things for oneself and one's family and community). For example, the first physical full-sized car you own may be very useful and fun, but the 10th such car you own probably won't have nearly as much value to you. A study discussed a couple years ago on Slashdot suggested that in the USA money was closely correlated with happiness -- but only up to around US$75,000 a year. After that, more money did not make much of a difference.
Further, there is even at some point *negative* returns for more goods and services, where the addition costs more of our time and emotion than it is worth. Suniya Luthar has researched problems affluent families have. One is disconnection from community. For example, poorer denser communities may have face-to-face activities like regular barbecues with neighbors, whereas wealthy people living on big plots tend not to interact as much with neighbors. See:
"The Culture of Affluence: Psychological Costs of Material Wealth"
Or, although it mostly relates to problems from a culture of competitiveness:
"In a surprising switch, the offspring of the affluent today are more distressed than other youth. They show disturbingly high rates of substance use, depression, anxiety, eating disorders, cheating, and stealing. It gives a whole new meaning to having it all."
Another aspect of affluence is that everything we own also owns us because it has to be maintained and stored and worried about. Many religions talk about a the value of a life of voluntary simplicity to make emotional room for other things that matter. Again, maintaining one car may be a chore but worth it, but maintaining close to 1000 vehicles like Jay Leno does means you are essentially running a private car museum as a full-time job. Or, you have hired someone to do it for you and manage that person (although I guess in the future you might have a robot maintain all the cars). If you like running a car museum, great, and it is good that some few people do such things. It is even better if they do so as a public museum. And to some extent through his videos, that is what Jay Leno does. And he no doubt gets a lot of satisfaction from that. Running a car museum though is a very difference activity than owning a car for basic transportation. And for most people, unless the really love cars, owning 1000 real cars would be considered a hoarding pathology. It would greatly diminish that person's quality of life by taking up time probably more healthfully spent in a variety of other ways. Also, while the USA today probably could benefit from more good car museums (we joined our local one last year to support it), at some point, with too many of such museums, each museum won't have much of an audience. Granted, issues about how individuals want to spend their time and resources versus their responsibilities to family, friends, neighbors, and the rest of society is a complex topic. But if some people need cars (and food) and don't have them because they can't afford them because they can't get a job and otherwise have no income, and other people are running private car museums full of 1000s of cars they never use, there will be great social stress -- even if the car museum owners are willing to employ the carless for US$1 an hour to dust the cars instead of installing HEPA filters in the museum and buying car dusting robots.
That said, my own kid probably owns 1000+ vehicles if I counted toy cars, LEGO models, and cars acquired in racing games or simulations like Gran Turismo, Rigs of Rods, MineCraft, and so on. :-) However, there is also a bag of eight unopened Hot Wheels cars that has been on the kitchen table for several days from a recent trip to the store -- an example of the point about the law of diminishing returns. Or even negative returns, because it would be nice to use the kitchen table for other things. :-) My kid has been recently building space craft in "Space Engineers" while those cars unopened. My kid previously built vehicles in Minecraft, and before that, using LEGO. That illustrates another point related to Maslow's Hierarchy of Needs. The needs higher up the pyramid tend to be more about self-actualization and doing. So, making cars or space craft may be more enjoyable than getting them pre-made (if you like that sort of stuff). In general, a lot of the self-actualization needs don't require that many supporting goods and services because they may entail creating stuff from raw materials for a sense of personal satisfaction by doing. The ability to enjoy virtual creations (which take less resources) rather than demand physical creations also does not bode well for creating a future demand for lots of new jobs making stuff (even as it may be good for the environment). Again, my kid and other children my kid plays with are willing to make stuff in these simulations without pay -- just because it is fun or meaningful to them. Dan Pink talks about how material reward is not a good motivator for creative efforts anyways, suggesting how an economy mostly based around though stuff and perhaps artificial scarcity may well have different rules than one based mostly on physical stuff and real scarcity.
"RSA Animate - Drive: The surprising truth about what motivates us "
There are other deeper issues. A finite Earth has only so much room for stuff, especially if we want to preserve ecosystem museums and biodiversity by letting big parts of the planet remain wilderness. So, we either need to limit production, meet those needs by virtual reality (like for my kid and cars), or we need to expand into outer space. Space is a great and noble challenge, and yes a source of some new jobs -- except space will be even more the domain of robots given environmental challenges and long transit times. Also, once we all have a lot of stuff, having more stuff (like Jay Leno has with 886 vehicles) loses much of its status value in our culture. See James. P. Hogan's 1982 novel "Voyage From Yesteryear" as a good sci-fi example of a culture of scarcity clashing with a culture of abundance. If stuff is cheap, then you might just as well have public libraries of cars (or rental agencies) where you can check one out to try it out and return it when you get bored with it. That greatly reduces the overall demand for stuff, since the production, maintenance, and storage of it is shared. We've seen that with computers, where we moved from mainframes maintained by experts, to personal computers maintained (poorly) by individuals, and now back to the cloud of shared computing maintained by a relatively few people.
Still, for all that, because it hinges on controversial expectations of "human nature", the issue of whether demand is limited for goods and services in the next few decades is IMHO much more debatable than whether most humans can compete economically with smarter machines. "The Ballad of John Henry" suggests what happens when humans compete with machines for physical labor. Deep Blue for chess and Watson for Jeopardy suggest what the result may ultimately be when humans try to compete with thinking machines. But if either mainstream economics assumption ultimately proves false, then we will have fewer human jobs as automation gets better (short of "make work" jobs like in Kurt Vonnegut's "Player Piano" novel) and we will have a huge social crisis in the USA (requiring interventions like a basic income or other big political changes).
To be clear, I think there will always be things for humans to do. Being a good parent, informed citizen, helpful neighbor, or supportive friend can generally take as much time as we have and then some. And people generally like to do things just because (like hobbies). The issue is whether most humans will employable down the road in a capitalistic marketplace compared to robots, AIs, and automation -- or as was suggested in 1964 in The Triple Revolution Memorandum, maybe we are finally reaching the point where: "The continuance of the income-through jobs link as the only major mechanism for distributing effective demand -- for granting the right to consume -- now acts as the main brake on the almost unlimited capacity of a cybernated productive system."
Although maybe putting a brake on endless production is actually sometimes a good thing as long as capitalism still too often ignores externalities like pollution? :-) Or when humans may tend towards addiction given their preferences are adapted for a certain level of material scarcity and our health is adapted for a certain level of exercise? On the other hand, if humans were less stressed out about work and competitiveness and financial worries about precarity, maybe they would not be so prone to addiction?
(BTW, I'm the article submitter, and I appreciate the great discussion here. I'd agree there is lots of room for uncertainty here, especially in the short term. I'm glad to see more and more people understanding the issue in my sig that the greatest challenge of the 21st century issue is the irony of technologies of abundance in the hands of those still thinking in terms of scarcity.)