Summary

In this whitepaper I will use trends from automation and video game revenue models to make the following predictions about the future of low-skill work.

In broad outlines, the argument is that automation and other trends will dramatically increase inequality in technologically advanced countries. Low-skill workers will find themselves largely unemployable in most of their traditional occupations. Meanwhile, ongoing wealth gains at the top of society will increase the ability of privileged people to devote time and money to leisure activities, especially them video games. Big spenders already support the game industry’s current free-to-play revenue model: It only takes a few high-spending players to offset the free access given to anyone who wants to play. The model works because the free players, those who spend no money, make the high-spenders feel better: Their presence in the game creates a real human community. The presence of other people is valuable to the big spenders for several reasons: The thrill of being richer and more powerful than others, yes, but also the simple happiness of companionship. Whatever the reason, big spenders are willing to pay a company enough money that the company can afford to open the doors to anyone. It is a small step from this model to one in which the game company directly pays the underlings for their presence. For the company, the added cost from paying players to come into the game and be relatively poor and underprivileged will be exceeded by the added revenue from big spenders who will enjoy a larger virtual society to befriend (or dominate). The number of players for whom this kind of model makes sense will increase as the 21st century unfolds. The number of real-world low-skill unemployed will increase, and the incomes of the high-skill technocratic elite will rise. In other words, both extremes of the play-for-hire model will grow as automation increases. Within a generation, playing games for money will come to be seen as a legitimate occupational choice for those whose skills are not valued by brick and mortar labor markets. Play-for-hire will become a primary way that the income gains from technological progress will be distributed to the low-skill population.

1 The coming wave of inequality

The technological system within which we live has irresistible imperatives to promote the development of more efficient means of production. Intelligent systems are increasingly deployed to perform work that low-skill humans used to do. At the moment I write, people are no longer employed to handle money, and those who drive vehicles for pay are in the cross-hairs of automated driving systems. It is apparent that most jobs that require little skill will eventually be done by a digital entity.

In the past, the invention of new machines for doing low-skill work caused disruption, but also led to an explosion of new employment opportunities for low-skill workers. In the industrial revolution, automated looms put weavers out of work, but they also created a factory system whose explosive growth provided employment and wages for millions of low-skill workers. It is often overlooked that the industrial revolution caused a population explosion of unprecedented magnitude. That would not have been possible if low-skill people were unable to find work that would feed many new mouths.

When a technological change puts millions of people out of work, a social change must necessarily result, one in which some work becomes available. With the factory system, the work that did become available was dirty, oppressive, harsh, and crushing to the spirit. But it was work. People survived. They survived because, having lost their incomes from farming or weaving or whatever old craft they used to do, they sought about for some way to make money, some way to eat. In the 19th century, they found their way to factories. And so, populations migrated from countryside to towns, cities exploded in size, the urban proletariat was born, and so on and so forth. Automation put people out of work, but those people went off and did something else.

In our current moment we must ask, where will all the drivers go? Forrester predicts that automation will cause a net loss of 7 percent of current jobs over the next decade. That’s just the beginning. If we think about the low-skill people who are about to be displaced, we can wonder, what sort of work will they be able to find? Or, be driven into? Their children will come of age knowing full well that they cannot live by the usual jobs of low-skill people. By what means will these new generations of low-skill workers gather up enough food and housing to stay alive? For if there are any means at all, they will surely seek them and secure them. They will find something to do, something they can do that machines cannot, and that others who have means will pay them for.

Who are those others, the ones with means? They are the people with high skills. As the digital age progresses, there will continue to be high demand for the services of bright people, the creative ones, the emotionally intelligent ones, the best of the best human servants. Entertainers, technicians, engineers, performers of all stripes, will do better and better. Anyone whose special gifts are impossible for a machine to recreate, will find themselves doing very well indeed. And then the other group who will do well, of course, are those who own the machines themselves. Those who own a share of the profits produced by machines will gain income at the same rate that the machines displace low-skill workers. Every time a self-driving car puts a taxi driver out of work, the owner of the car earns more money. Thus there will be three types of wealthy people: Those who own machines, those who design / build / operate machines, and those who provide services that only people can provide.

The system as a whole will get richer and richer. Every time a machine replaces a person, it does so because it is better at the job. It produces more at the same cost, or the same amount at a lower cost. Otherwise, there would be no reason to make the replacement. But this means that each time a machine does a person’s job, the economy as a whole gets richer. More stuff is made at lower cost. Automation is a force for efficiency. It induces economic growth. It makes money. And that money will go to the people at the top, the people with ownership or irreplaceable skills.

Automation thus necessarily induces inequality, at least in the first moment. One man is out of a job, another man makes more money on his assets. But these unequal slices are part of a growing pie. The question becomes, what happens in the second instance - after the driver loses his job, after the car-owner gains more wealth. What happens then? What new social arrangement might happen? Is there some new way that the rich person can hire the poor person? A new system whereby the money of the wealthy moves into the pockets of those who are looking for anything to do in return for money. If the car driver can no longer provide car-driving, can he provide something else?

In the very near future, we will be facing the reality of hordes of low-skill workers with no meaningful work. Indeed, this is already happening. In the late 1970s, 7 percent of men in their 20s with less than a bachelor’s degree did no work at all in the preceding year. In 2015, the number was three times higher, 22 percent. Unpublished research by economists suggests that large numbers of low education young men have abandoned the world of work simply because of the the joys of video game play. Facing a choice between seeking work or staying at home playing games, many young, low-skill men seem to be choosing the games. It is a sign that the real world of work is becoming less rewarding for those with low skills.

2 Options for the low skilled

In the 19th century, low-skill workers went to work in factories. What will they do in the 21st century?

One common idea is that education can change these dynamics. It can’t. It is patently unfair to try to address this problem by pressuring people to go to engineering school. For surely all this will do is start an arms-race among high-skilled workers. If the technological economic system requires only 10 percent of the workforce to operate at full capacity, then only 10 percent of the workforce will be employed as operators. Training up millions more people to have the necessary operating skills will only unleash a vicious competition among them, driving down the wages of those who do get the jobs. Consider: If you have 100 people whose skill is ranked from 0 points to 100 points of ability, and you will hire only the top 10, it does not help anybody to teach them all 5 more points of skill. You will still only hire the top ten, the people with skills from 95 points to 105 points.

Skilling up the workforce only makes sense if there will be an increased need for high-skill people. But the dynamic of technological explosion in which we live is such that for every new high-tech job that is created, two lower-tech jobs will be destroyed. This is simply because people are more expensive than machines. The system will do everything in its power to get its work done with the minimum number of people. It is going to try to shed people whenever and wherever it can. Teaching low-skill people to be somewhat better technicians than they currently are is a losing strategy for them.

Instead, we must think of other ways that a low-skill person can do paid work for others. What kinds of things can low-skill people do that machines simply cannot? This question forces us to unpack the concept of “low skill.” What we really mean, by now, is that low-skill people are the ones that machines can replace. They are currently working in a way that a machine can do. Whatever is special and unique about human thinking, whatever we have that machines cannot have, may well be present in the minds of the “low skill” workers; but right now, that special sauce is not a part of their work. “Low skill” workers may have high skills, at something; but the work they do today is work that a machine will be doing soon enough.

If “low skill” workers are becoming defined as “those whose jobs will be automated,” it follows that if those people are to survive, are to eat, they must bring up within themselves, and express to the world, some set of skills that cannot be automated. It also follows that the term “low skill” applies to a lot more of us than would have been expected under traditional definitions of the term.

Earlier I argued that the only people who will be compensated in the automated future are a) those who own machines, b) those who operate machines, and c) those who perform services that only people can perform. If the “low-skill” displaced masses do not own machines and are unable to become operators, only one avenue of compensation is open to them: Service.

The question is, what kind of human-only services can low-skilled people provide, in massive numbers? For there will surely be huge masses of displaced low-skill people. The answer here is not going to be tennis instruction or lawn care. In the future we are facing, if every rich person had 10,000 butlers and maids, it would still not be enough employment. The technological disruption will therefore drive some kind of huge social disruption, that creates new ways for a few privileged people to move income toward massive numbers of the technologically unemployable population.

3 “Low skill” people can play games

All of us can play video games, regardless of skill. The game industry excels at creating systems in which both angry lawyers and stoned teenagers can make their way. Games offer a range of skill challenges, by design. Games are designed so that all the players have fun, including those whom the outside world would call “low skill.”

Skill effects within games are tightly managed by the designers. Designers make acutely conscious choices about when, where, and how the real-life cognitive and physical skills of the players will have an influence on game outcomes. Games are now designed to adjust dynamically and automatically to the perceived abilities of the players. Is this player dying quite a lot? Remove some of the zombies. Not dying enough? More zombies! It could be argued that games are designed to reward skills that the market finds less valuable, precisely in order to capitalize on the way people feel underappreciated. The point is that whatever the outside world may think of skills, game design seeks ways to make every player appreciated, regardless of skills.

Designers also carefully manage when, where, and how the real-life monetary resources of the players can affect the game. In many games, it is possible to spend extra money to get a better sword, or more life potions, or unlock a faster horse or a new level. Whereas in other games, no amount of outside money can affect what you can achieve in the game.  

Designers manage skill, money, and time as well: Some players have quite a lot of time to spend in a game, others don’t. It is up to the designers how long it takes to achieve things in the game.

As a result of these designer decisions, the game industry finds itself unwittingly serving as a vast global agora of skills, time, and money. Designers can increase or decrease the impact of player skill, money, and time input, and in so doing they give their games a certain profile. Different games appeal to different players. Players can move from game to game, according to their resources and tastes.

Some games are so big in scope, time length, and space that they can accommodate widely different types of players. Different parts of a single game may appeal to the skill-rich, the money-rich, and the time-rich. This is seen most clearly in the free-to-play revenue model. A game run on the F2P model opens its doors to anyone. Anyone can play the game, free of charge. At some point, however, some aspect of the game either requires or encourages payment. Perhaps the game has 100 levels, and the first 50 are free. You have to pay to unlock the other 50. Or the game has 100 levels, all completely free, but it takes two years to get to level 100, unless the player buys some sort of special equipment. F2P games make money by charging for extra game features.

The free to play model has exploded across the industry in the last decade. It has turned out to be fabulously lucrative. A free game has the lowest possible barriers to entry, and can quickly gain a large population of players. Those players create buzz and excitement; they confirm to the world that it is a good game. Production costs are such that even if only a fraction of players ever pay for something, the revenues gained far exceed the costs of providing the game free to all the others.

It is understood colloquially among game designers that the revenues of F2P games follow the same patterns as casinos, in that a few big spenders are enough to make the casino turn a profit - even if you give free drinks to hundreds of low-spending people. These big spenders are known as “whales.” The game industry has whales as well. There are people who spend thousands of dollars every month on the most trivial of game items. A very large portion of revenues are provided by a small percentage of the player base. Gamasutra reports that less than one-fourth of one percent of F2P players generate almost half the revenue. “Conversion rates” - the percentage of players who spend any money at all - can be as low as 3 percent or 1 percent, yet the game still turns a profit. The free spending of a few is sufficient to support the game.

What then is the role of all the other players? What purpose is served by all those people in the casino who drink their free drinks and gamble away comparatively tiny amounts of money? Those people are a critical part of the revenue model, for they form the social environment within which the whale can make friends, and compared to whom the whale looks like the awesome person he wishes to be. Without the free-drink gamblers, the casino whale has nobody to talk to and nobody to impress with his casino privileges and private rooms. The same holds for F2P games: Without the free gamers, the spending gamer has nobody to talk to and nobody to impress.

Thus there are two elements that explain why big spenders will provide revenues to companies that alow free access: A communion aspect and a comparison aspect. The communion aspect is easily enough understood: Rich people get lonely too, and a thriving game community gives them a social world for making friends. The comparison aspect is a little more complicated, so let’s unpack the social dynamics of a F2P game. Consider three ways that a person might spend real money.

Note that each of these ways of spending enhances a comparative difference between the spending player and everybody else. Cosmetic spend (1) enhances the social prestige of the spender, just as flashy diamonds do in the real world. Convenience spend (2) allows people who are short on time, or who are bored, to move quickly to more entertaining parts of the game. They get ahead quicker and look more competent with respect to the game systems. Competitive spend (3) allows people to dominate other people, to beat them, to get ahead. In all of these cases, the spender looks better than everyone else. Compared to people who don’t spend, he has better looking items, he skips the annoying parts, and he directly puts a beat-down on them. The spender is playing the game in first class.

Psychologically and socially, first class only exists if there is a second class. Thus the comparison role of all the free players is simply to be the second class. Their job is to sit in second class precisely so that the first class passengers can feel good about having been able to board first, get better food, and have more space. In an airplane, there are physical reasons why not everyone can have the privileges of the first class passengers. But in a game, there is no reason at all. There is no reason whatsoever why a game company could not give all players the same privileges they give to spenders. Thus the lines that are drawn between the spenders and the non-spenders are wholly artificial. They exist only because drawing them induces the spenders to spend. And therefore, one of the main purposes of the free players is simply to be present in the environment as those-who-did-not-spend. Their job is to form a comparison group, against which the spender looks impressive indeed. And then, alongside that comparative context, to be available for friendship, camaraderie, or romance. The free players make it a real social world, one worth paying for.

4 Playing games for money

Free to play games illustrate that there is a role in games for people who do nothing but play. They don’t have to be smart, they don’t have to be good, they don’t have to be pretty, they don’t have to be dedicated, honest, or loving. They just have to hang around, and - this is important - be PEOPLE. The humanness of the members of a game society is essential to the formation of meaning in that society. Only human beings can serve. It is a job only people can do.  

True, a savvy game operator could try to skunk his whales by populating his game with life-like AI. But word would get around, and one whiff of people-faking would drive all the whales out overnight. The game Barbie Online had troubles because no real boys were signing up, so the company added artificial boys. It didn’t work; players knew the boys weren’t real and the game fizzled. The same thing happened with Canadian dating site Ashley Madison: Many of the “women” on the site were bots, but it only took a few email leaks to bring the scam to light. It is one thing for one computer to trick one person in one interaction - the “Turing test.” It will be a long, long time before a crowd of AIs can fool a crowd of people for years. It will be strange times, though, as game companies will guarantee sentient players the way grocery stores guarantee fresh vegetables. On the other hand, nothing kills a game like empty servers. Locations are designed to keep the real people bumping up against one another. Critical mass is critically important.

In games, we already see how valuable pure humanity is. Whatever the markets may say, pure simple humanity is a good thing. People with means are willing to pay money to be in community with other folks who get in for free. But how much will they pay? Is the willingness of big spenders to spend enough to send the street cost of game-playing into negative territory? Consider: Twenty years ago, all players paid the company for the right to play the game. These payments fell and fell. Now they are at zero across wide swathes of the industry. Why shouldn’t the price trajectory continue into negative territory? Can the spenders support not only the game-playing of poorer people, but their incomes as well?

Consider again the F2P revenue model. The iron equation of FTP is eLTV > eCPA + Ops, where eLTV is the expected long-run revenue value of a player, eCPA is the cost of acquiring that player, and Ops is the operations cost of the game. Generally, what seems to happen is that large numbers of free players are attracted into the game initially, but then most of them fall away. Over time, the paying players make up an increasing share of the player base. As this happens, companies spend less on acquisition.

What does “acquisition” mean? Typically, this is a marketing cost. Advertising - on mainstream TV, for example - can drive millions of players into a game. So long as the marketing costs (plus operations) are exceeded by the revenues that these players bring in, the marketing is worth doing.

We are considering here a second possibility, one that has not yet been tried but is surely around the corner: Paying players to play. If a large player base is important for a game, and if the main source of revenue comes from a comparatively small segment of the player base, it might well make sense for companies to pay players to come in. Paying players directly is just another acquisition cost. Instead of marketing to people, you pay them.

Initially, paying players will look like the lame maneuver by developers who lack confidence in their product. But the same was once said of developers who gave their games away for free. The stigma against F2P fell when the industry realized how much money it made. One can expect a similar acceptance of play-for-hire.

Some games are doing this already, though indirectly. In EVE Online, it has become possible for a player to earn enough in-game currency to purchase a pass that allows them to play for free for one month. The pass card replaces what would have been a real-money cash payment. Since the company sets the rate at which in-game currency can be earned, it is effectively paying the players for their in-game labor, which it then allows them to translate into a good - the pass - that has real-world cash value. Another game, World of Warcraft, has a similar system. Indirectly, these games are hiring people to play.

And why not? Economics teaches that transactions should occur whenever they are beneficial to both parties. Companies will almost surely gain from paying players; it is probably more efficient than marketing at getting people into the game. As for the player, it is surely the case that for some people, earning money in a game is just better (almost as lucrative and much more enjoyable) than earning it out of game. The in-game compensation is better than the out-of-game compensation. It is a better job, all things considered: For for-hire players, one suspects, the money would be just so-so, but the work environment would be outstanding.

This applies especially to the situation of a low skill worker who has just lost a job to automation. His choices are to take a worse job (that he will surely lose in the near future), try to skill up, or go begging. What person facing this situation would not leap at the chance to earn money while playing a video game? Thus it stands to reason that the number of people willing to work in games, even for small amounts, will increase as automation does. Paying players to pay generally makes sense for the game companies, and will make sense for more and more workers as time goes on. It is a practice destined to grow.

The typical dynamics of player bases - with the free players coming early but then leaving, while whales stick around - suggest that the practice of paying players to play games will be concentrated at the launch of a game. The company will allocate budget to the hiring of millions of players, anticipating that they will be in the game for a relatively short period. Over time, these player wage costs will fall. Yet while the for-hire players are there, the whales will become dedicated to the game for the long haul. The spending of the whales will then sustain the game for a long period. As for the for-hire players, their lifestyle will be one of migrating from game to game, picking up rewards during the launch phase of new games and then moving on when the money starts to dry up.

5 How play for hire will emerge

Right now, the environment for compensated play is clouded by overlapping institutions and regulations. Games often have multiple currency systems in place. The real world does too, of course. Laws and regulations constrain this space in ways that nobody yet understands. Yet some things are clear.

The practice of working for money online has become widespread if not yet common. New research indicates that large numbers of people around the world participate in a global market for online jobs. Oxford University has recently launched the world’s first Online Labour Index. The online labor market exists and seems to be growing.

Precedent also exists for the idea that wealthy people can pay extra money to make a game more enjoyable. Games have their shops, and game companies accept credit cards, debit cards, Paypal, and various kinds of points and credits and time cards and what have you. There should be no problem sending purchasing power from the vaults of the rich into the vaults of the game developers.

It is the ability of poorer people to get money out of games that faces legal questions. In principle, a game company could simply hire players as employees. This is highly unlikely to be a first step (or a second or a third or a 17th), however, since the regulatory overhead on hiring is incredibly burdensome. What’s more likely is a system that evolves from payment practices that already exist.

Under current US law, anyone who makes money in a game has to report the earnings when they are liquidated in the form of real-world currency. It is not clear what would happen if someone were to earn gold inside a game and then trade that gold for something in-kind, like rent. But this could happen. It could happen today, in fact. A player can already earn in-game money and give it to his landlord in lieu of rent. Players can already sell their game assets for real-world cash. Thus there is already some precedent for players to be able to transfer some of their in-game earnings into outside purchasing power.

Also it is important to realize that the world of finance is facing a wave of disintermediation. Technology is beginning to allow an almost unlimited variety of ways to transfer purchasing power among people. Technology also allows game companies almost unlimited ways to craft incentives without falling afoul of the law. For example, today, one would not want to allow whales to directly hire underlings within a game for anything that could be liquidated, because this would enable money laundering. But there is no opportunity for money laundering if whales pay $1 million to the game company and the game company provides one million piles of random, sellable loot worth $1 on the eBay market.

There might evolve a market that translates game time into game assets, such as time cards, that can then be quickly and easily traded for real money on the outside. Or a game company could make a deal with a credit card company so that game time yields bonus points on the card. Since the card can be used to pay rent and buy groceries, the game time translates quickly into real purchasing power. Current law covering these points systems does not handle them as income, so there would not apparently be any tax or reporting burden. Another possibility would be to pay players with an undercover currency, such as Bitcoin. These types of changes will represent a minimal disruption to current practices, but they will be the first signs of an emerging general change in employment patterns.

Once the basic idea is established of paying players to play, there will be some sort of legal and regulatory push to normalize the practice and make it explicit. There will be controversy. Initial objections will come from different parties. Developers, or at least the purists among them, will say that paying players to play undermines the whole idea of a game. Perhaps it does. But since it makes economic sense, it will happen anyway. Pundits and lawmakers will say that players are being exploited, that games are becoming sweatshops. Perhaps. But as inequality reaches soul-crushing levels, sweatshop employment within a game will look like a fairly humane alternative for people who cannot find work. Academics will lament the transfer of inequality and oppression from the real world into the game world. Yes, that will be sad. But since it will allow some people to eat who otherwise would starve, it will have to be accepted. The transition will be no different from that which drove farm hands into the factories. Sad; much was lost; but it was the best alternative. Social critique, policy, and aesthetics will have to give way. That’s why it is called disruption and not merely change.

After the play-for-hire model has been normalized, low-skill work will shift into games on a large scale. It will be found that for most people who do not have a particular skill or an inherited lump of wealth, one of the best ways to spend time is to play games and, with the meagre earnings, make the best life one can. Consider the minimal lifestyle necessary to sustain a full-time job playing games:

Given that the work can be performed anywhere with internet, the costs of space and food can be very low. The needs of a working gamer will cost no more than a few dollars a day. Although, to us, this looks like a terrible lifestyle indeed, remember that this person is spending all of his time inside a game, a place designed to make him feel quite good. True, there will be other people in that space who are doing even better. The people with money to spend in there will do so, and they will have the most powerful weapons, the fastest cars, and the best-looking outfits. Yet the in-game difference between the rich and the poor can be managed in ways that are impossible out-of-game. The wealthy people in the game can be shown how very wealthy they are, while the poverty of the poor can be hidden from the poor themselves. The comparisons of rich and poor, and the emotions and self-esteem of them all, can be managed by the game developers. Thus while the outer world will provide nothing but glaring contrasts, the in-game world will be everything to everyone.

As developer Gordon Walton once said, “Everybody wants to be a hero.” In the play-for-hire game, everyone will still be a hero. But some of them will be paying a lot of real money to be a truly amazing hero, while others will be receiving a very modest amount of real money to be a pretty good hero. They will all be happy. And by this means, the vast wealth produced by automation will trickle down from the owners and the operators to the “low-skilled.” For it will turn out the that low-skilled were not all that low-skilled after all. They all have something that every human has, an ability to entertain other humans, just by being human. Every player of a game provides entertainment for every other player. It stands to reason that some people, even many people, could get paid to do this entertaining. And the job of being an in-game entertainer will become ever more attractive as the jobs available in the outside world steadily decay.

6 A timeline

The developments discussed in this whitepaper are already underway. I am told that developers have been aware of the option to pay players but have held back because of legal and regulatory barriers. And yet, as we have seen, young low-education people are already exiting the workforce in much larger numbers. Games are already taking advantage of distinctions between time-rich, skill-rich, and money-rich players. The net cost of playing games is becoming negative for some players.

Within five years, these trends will result in the implementation of a new business model in the game industry. At least one game will innovate around current regulations and set up a systems in which most players are given financial incentives to play. The incentives will be indirect: Points on a card, virtual items that can be sold for money, etc. These will be known at first in the industry as “player-retention incentives.” They will probably enter discussions as a way to lower churn in the player base, to keep lukewarm players around who might otherwise leave. The game will have a strong pay-to-win element, and will earn huge revenues from spenders. Compared to these revenues, the costs of “player retention incentives” will seem trivial.

The next game in the cycle, still within five years, will go a step further and offer “player retention incentives” to all players from the beginning. This model will also succeed, and the industry will become aware of the new model. After that, most new games will try to develop innovative ways of paying their players.

Within ten years, these innovations will be complete. Game companies will have created streamlined and efficient ways to channel purchasing power from richer players to poorer ones. A share of the revenues provided by whales will be spent efficiently to generate launch buzz and possibly to keep the overall player numbers up. This will be a standard tactic in the industry.

At the same time, in the outside world, automation will continue. Job prospects for low skill workers will keeping eroding until it reaches the point that the reservation wage - the minimum wage for which a worker is willing to take a job - will have fallen well below the legal minimum wage. This means that low-skill workers will have greatly increased their willingness to work in games. For many young people, the fantastic work conditions will more than offset the low pay that games offer. Working in games will be more attractive than any job available in the real labor market.

Nonetheless, at this point the game workers will still not generally see themselves as such. They will see themselves as unemployed and unemployable. Yet they will say that they are keeping themselves afloat by going into games and “playing smart,” bringing out whatever purchasing power they can. It will at first go almost completely unnoticed that that purchasing power is becoming large enough to cover the meagre costs of living which most low-skill workers have come to see as normal. Low skill people will have become accustomed to living far away from urban centers, with no car, no entertainment, minimal clothing, basic housing, no health care, and poor food. Their lives will be lived on the internet. And on the internet, through pay-for-hire games, they will earn enough to support their minimal bodily needs.

Within 20 years, these trends will have changed the relationship of low-skill people to the labor market. Among those who have no technical abilities and no ownership shares of capital, many if not most will earn their living within multi-player online environments. There will be millions of play-for-hire players all across the globe. If populations in China and India participate, the numbers will be in the billions. For-hire players will have become conscious of this social role; they will understand that playing is their contribution to society. They will express pride at the quality of their play. Indeed, those who do not “pay to win” will be quite proud of their hard-earned accomplishments relative to the wealthy who buy their way forward. Players will see themselves as experts in their game of choice and will think of that virtual place as an important social and even political entity. Players will be organized into pressure groups. There will be uprisings and strikes and boycotts in games where compensation is deemed unfair. Meanwhile the wealthy will jump from game to game, always seeking the very best experience for their money. The sad history of inequality and politics will gain a number of new chapters, this time online.

Of course, there are a number of trends that go in a different direction. Fertility is falling in the developed world, which undercuts any labor-surplus argument. In the developing world, incomes are rising, which should increase demand for goods from developed countries. Half the population - women - do not seem to fall into intensive game play in the way this argument assumes. Any prediction could be wrong.

Yet there is a trump card here: The argument here is ultimately driven by technologies of AI, robotics, and virtual reality, and technology, at this state of evolution, is beginning to dominate every other trend. It is not unreasonable to assert that tech is becoming the only trend that matters. Technological change has become so powerful and rapid that it is now fair to say that tech does not react to society, tech creates society. That includes everything: Law, gender norms, employment, and, of course, games. Even climate change, that great problem of our time, will fade as a public issue as technology remakes how we live. How much energy is required to keep 5 billion unemployed people happy in their comfy, VR-enabled chairs?

If technology is truly a dominant force, then we can expect technologies of entertainment and automation to create a vast labor market for low-skill game players. It is a new service sector that will thrive in decades to come.

Acknowledgements: I am grateful to Mike Sellers, Joshua Fairfield, and Isaac Knowles for helpful comments.