The modern economy depends intrinsically on a bargain that may not survive automation.
Most people work. They earn money from the work. They use that money to buy what they need. Work is how society gets tasks done and it is also the main way income is distributed. The U.S. government likewise relies primarily on taxes tied to individual income. For most people, the right to survive in the economy is derived from their ability to sell their labor.
Artificial intelligence and robotics will eventually break that arrangement. We might question if every current claim about AI is true, whether any particular forecast will happen on schedule, or if some forecasts will ever happen at all. However, I expect that machines will eventually become better and cheaper than humans at most economically useful tasks. When it happens, what new economic relationship can we build to replace the old one?
When machines are capable of most economically useful tasks, unemployment will no longer mean what it meant in earlier economic crises. In the past, mass unemployment meant idle workers and reduced production. But when machines are doing the work, production can continue to rise even as human employment collapses. The problem will not be scarcity, at least not in the familiar sense. The problem will be distribution.
That distinction points to a distressing irony. A society full of robot factories, automated farms, AI doctors, AI lawyers, AI programmers, and AI teachers will be materially rich. It will produce more goods and services than any society in history. But unless people own the machines, control the platforms, or receive some share of the surplus, they will have no ordinary way to participate in that abundance.
Machines doing our work should make us wealthier and freer. The danger is that we keep treating employment as the price of access to the wealth they produce.
Universal Basic Income (UBI) is the most familiar answer. If machines produce the wealth, then the UBI solution is to tax the machines or their owners and send people checks. But tax-funded UBI assumes that governments will successfully execute a large and politically durable redistribution program under conditions of extreme technological and economic disruption. That may happen, and it may even be necessary. But it would be unwise to assume that it will be enough or that it will happen in a timely way.
There is also a deeper problem. A monthly check may solve the problem of survival, but it does not solve the problem of what people are supposed to do.
Work does more than provide income. It creates structure. It offers status, routines, goals, social roles, achievement, and outlets for ambition. Some people might thrive in a world of financial security and open-ended leisure. Many others would not. A society that replaces work only with checks risks solving the material problem while leaving the human problem untouched.
That is why society needs to consider a strange possibility: paid, structured human activity designed for a world where machines no longer require human labor for production. This is the Omni-Game economy.
Contents
Games That Pay All Players
The name is intentionally odd, but the idea is straightforward. An Omni-Game would be a large system of activities that pay people to participate. These activities would not be limited to video games, virtual reality, or entertainment. They could include physical contests, puzzles, exploration, sports, markets, creative challenges, social tasks, scavenger hunts, exercise, nature activities, and nearly any other form of structured human activity. What they would have in common is that they would be designed for humans, protected from machine competition, connected to real income, and not dependent on producing economically useful output. Some activities might produce incidental benefits, but payment would reward human participation rather than the market value of those benefits.
This sounds frivolous at first. But many systems already exist in which play, competition, or game-like activity produces income.
Players have long earned real money from virtual economies, from gold farming in online games to virtual property markets. Casual game platforms pay small amounts of cash or gift cards for engagement. Gambling, both online and in person, pays winners according to rules set by the game. Tournaments convert structured competition into prizes. Prediction markets let participants trade contracts on verifiable outcomes. Many financial markets, while tied in part to real economic activity, also have game-like properties: speculation, scoring, risk-taking, and collective belief.
The important point here is not that these systems are good models. In fact, most are bad models.
Nearly all existing paying games are designed to extract value from participants. Casinos come out ahead. Platforms take fees. Advertising-funded games pay players only because advertisers expect to recover more from the players than they give away. Even systems that seem to reward participation usually depend on most participants losing, spending, or being monetized in some way.
The Omni-Game would reverse that logic. Its purpose would not be to extract money from players. Its purpose would be to pay them.
There is a further distinction between a game that could make money for a player and a game that exists specifically to distribute money to players. Games in which a few participants make money already exist. What is proposed here is a system of games built specifically to distribute income broadly through human participation.
Why Games Rather Than Jobs?
A working Omni-Game would need to satisfy several conditions.
First, income from full-time participation, either on its own or in conjunction with UBI or similar support, would need to be enough to live reasonably. Otherwise the system would fail at its primary purpose.
Second, earnings would need to depend on some combination of time, effort, skill, persistence, and luck. If rewards were substantially disconnected from activity, the system would become only a disguised transfer payment, like UBI with needless extra steps. That may still be useful, but it would not replace the motivational structure of work. Humans are strongly motivated by the possibility that effort, skill, persistence, and even luck can matter.
Third, the activities would need to provide challenge, growth, and safe outlets for competition. Paying people to click meaningless buttons would be little better than make-work and would have the same deadening effect as pointless bureaucratic tasks.
Fourth, payment must not depend on producing economically useful output. The system would reward human participation itself, regardless of whether the activity produces anything that someone else would pay for.
This combination is the defining design constraint: activities that provide a livable income and are psychologically and socially meaningful, detached from economically useful output.
This is why they are games rather than jobs. A job exists because someone wants the output. If the output has economic value, then there will be an incentive to automate it. When robots can build houses, cook meals, diagnose diseases, write software, teach classes, or compose music better and cheaper than humans, most people performing those jobs will be unable to compete. A market that rewards output will reward the machine. Once output has intrinsic economic value, barring machines from the work would simply be legislating inefficiency, and the inefficiency would inevitably be resisted by the people paying for the output.
An Omni-Game avoids that problem because payment rewards human participation rather than useful output. There is no incentive for the people controlling the system to automate the activity, and reserving participation for humans does not make resource production less efficient. They would gain nothing by letting bots slay dragons on its servers, run its scavenger hunts, trade against each other in its markets, or complete its puzzles. Players could still try to cheat and automate their participation to collect payments, but excluding those bots would not make the production of goods and services less efficient. In fact, excluding bots would be one of its core technical and institutional requirements.
That makes it very different from professional sports, streaming, influencing, or entertainment. A basketball player is paid because spectators value the performance. A streamer is paid because an audience watches. An actor is paid because the performance produces something others want to consume. Those may be enjoyable forms of labor, but they are still labor. They are still vulnerable to competition from synthetic performers, AI-generated entertainment, and machine-optimized attention systems.
The Omni-Game is different because the human activity is intrinsically valued, rather than being valued only as a means to produce an economically valuable output. The human activity is the point.
This does not mean that ordinary work would disappear entirely. Some people would still make things other people value precisely because they were made by humans. Handmade objects, human-generated art, teaching, caregiving, performance, craft, and entrepreneurship may retain markets. Some people will always find ways to create value. But realistically, those opportunities would be limited, and they would exist alongside machine competition. The Omni-Game would be isolated from machine competition. Anyone could use it as a structured way to earn income without having to outperform machines.
What an Omni-Game Could Be
Picture not one giant video game, but a broad and expanding network of human activities connected by a common payment system. A participant might earn through a chess tournament, a neighborhood challenge, a hiking route, a language puzzle, a team sport, a virtual expedition, a market simulation, or a creative contest. Some activities would be competitive, while others would be cooperative or solitary. Some would require intense skill, while others would reward patience, consistency, or exploration.
The games could be digital, physical, or mixed. They could happen in homes, parks, schools, community centers, and purpose-built spaces. A person who dislikes fantasy games might prefer gardening challenges or long-distance walking. Someone who cannot participate physically might choose puzzles, social strategy, or virtual exploration. A successful system would need enough variety that people could choose activities compatible with their abilities, interests, and temperaments.
Someone relying on the Omni-Game for income would still need to participate to pay the bills. It would free people from having to produce something someone else wants to buy, but not from every obligation associated with earning a living. Its appeal would depend on a broad set of options such that most people could find choices well aligned with their interests and skills.
The payout design would matter as much as the activities. There would need to be a reliable floor so that ordinary participation could support an ordinary life. Above that floor, variable rewards could recognize skill, effort, persistence, or luck. The system should permit ambition without requiring a permanent class of losers. Existing games usually create excitement by concentrating rewards at the top. An Omni-Game would need to create excitement while remaining a net payer to the population as a whole.
Modern game designers already wrestle with fairness, cheating, accessibility, addiction, ranking, moderation, and participant retention. Many successful games show that workable balances are possible. An Omni-Game would apply that accumulated experience to a different purpose.
Some activities might incidentally produce useful results. A walking game might improve a player’s health. A neighborhood exploration game might identify broken streetlights for a city. A nature challenge might generate observations useful to researchers. Those benefits would be welcome. But they could not be the economic reason for the payment. The moment the system pays primarily for useful output, the activity becomes work again, and the pressure to automate returns.
Where the Money Comes From
The obvious objection is sustainability. If the system pays people more than it takes from them, where does the money come from?
The answer is that it must come from the surplus generated by automation. In a world where machines perform most economically valuable work, the owners or operators of those machines will control enormous productive capacity. Robot factories, automated farms, AI service platforms, autonomous construction systems, and machine-run logistics networks will produce wealth at scales that make today’s labor economy look tiny by comparison. The Omni-Game would be a mechanism for distributing some of that surplus without the charade of pretending that humans would still be needed for production.
From one perspective, this looks bizarre. Machines produce goods and services. Humans play structured games. The people who own the machines give purchasing power to the players, who then use it to buy the goods and services the machines produce. Why would the machine owners short-circuit capitalism so that people no longer need to be the source of productive value in order to receive income?
When machines building machines are capable of nearly exponential increases in production, they will either be left idle or produce a growing surplus that has to go somewhere.
It can be hoarded by the owners of capital, redistributed by states, shared through broad ownership, or distributed through public and private systems that give people claims on machine-produced abundance. The Omni-Game is one possible form of that last category.
Power and Patronage
Why would anyone build such a system? One answer is power.
The sponsor’s return would be goodwill rather than net revenue extracted from players. A system that reliably pays people and improves their lives could earn loyalty, legitimacy, and influence. In an economy where machines make it easy to expand production, the support of millions of people could be more valuable than another increment of productive capacity.
That same relationship could create dependence and give sponsors power over participants’ lives. A badly designed Omni-Game could lead to a corporate welfare state, a ubiquitous surveillance platform, an addictive attention trap, or a softer version of the dystopias imagined in Ready Player One and Black Mirror.
But power is also one reason a socially positive, benevolent system might actually be built. Ultra-wealthy people and institutions often seek more than consumption. They seek legitimacy, recognition, influence, gratitude, and historical importance. They fund museums, hospitals, universities, foundations, and public works. Some do this from altruism, some from vanity, some from a sense of duty, and some from a desire for control. Usually it is a mixture.
An Omni-Game would be a public good of a very unusual kind: a way to provide income, structure, and social connection in an automated economy. It would not require pure altruism. It could serve the interests of its builders while also serving the needs of its participants. The goal is to align those interests in a mutually constructive way.
Competition from multiple systems could help, although it would not be sufficient. A single Omni-Game would carry obvious risks of domination. Several systems could offer different cultures, activities, rules, payout structures, and governance models, and people might move among them or participate in multiple ones. Governments could regulate them. Cooperatives could run some of them. Philanthropists, companies, religious organizations, universities, cities, or states might sponsor others.
A Post-Work Institution
Small versions could be tested before mass automation arrives. A city, university, cooperative, or foundation could sponsor a limited network of activities and study whether the system provides income, structure, connection, and well-being without becoming exploitative. These experiments could show how to balance baseline payments with variable rewards, how to verify human participation without invasive monitoring, and how people respond when an activity is explicitly meaningful to them but economically unnecessary.
Important questions remain about long-term funding, governance, and whether people would find participation rewarding.
Regardless, the logic has to be followed even when it leads somewhere that initially seems implausible. Frontier AI agents already perform tasks that recently appeared to require uniquely human judgment, and their capabilities continue to improve. My expectation remains that it is a question of when, not if, machines will outcompete most people in the job market.
That is frightening, but disliking the conclusion will not prevent it from arriving. Society will need an answer or bear the consequences. An economy can produce abundance and still fail the people who live inside it.
The deepest challenge of automation will not be teaching machines to do everything. It will be building institutions for a world in which productive necessity no longer organizes human life. When machines no longer require human labor, society will still need ways to distribute its surplus and reasons for people to get up in the morning. The Omni-Game is one attempt to imagine both at once.
Author Bio
James F. O’Brien, Ph.D., is a Professor of Computer Science at the University of California, Berkeley. His research interests include computer graphics and animation, simulation of physical systems, human perception, rendering, artificial intelligence, computer vision, virtual reality, digital privacy and security, and image and video forensics. He has also worked with film and game companies to integrate advanced simulation physics into games and special effects. His destruction-modeling methods have been used in more than 200 feature films and game titles. In 2015, the Academy of Motion Picture Arts and Sciences recognized that work with an Academy Award for Technical Achievement.
You can also find me on LinkedIn, Instagram, and at OBJF.ai.
Disclaimer
Any opinions expressed in this article are only those of the author as a private individual. Nothing in this article should be interpreted as a statement made in relation to the author’s professional position with any institution.
This article and all embedded images are Copyright 2026 by James F. O’Brien. This article was written by a human, and both an LLM and other humans were used for proofreading and editorial suggestions.
Image layout Copyright 2026 by James F. O’Brien. Collage: CC BY-SA 4.0. Adapted photos: Wolfgang Sauber, Eirik Irgens Johnsen/Kulturhistorisk Museum, Roy.wonder.cohen, Swathysuren/Hohum (CC BY-SA 4.0); Dolev, Panjigally/Hohum, Holger Ellgaard, Art Anderson, Ruben de Rijcke/Gwolf, Masur, Tiia Monto (CC BY-SA 3.0); Science History Institute, R. Henrik Nilsson (CC BY 4.0); Seth Morabito, Tony Webster, ajay_suresh (CC BY 2.0). Other photos: public domain/CC0.