Abstract
The Inequality We Might Want argues that humanity is entering a dangerous transitional period where AI will outperform humans across most productive domains long before societies are psychologically or structurally ready for a post-work world. Drawing on research in psychology, political science, economics, and technological history, the essay shows that humans depend on work for meaning, identity, structure, and mental health, and that abrupt attempts to eliminate hierarchy have historically produced collapse rather than equality. It introduces the concept of a computational currency: a merit-based, non-transferable, infrastructure-backed system designed to preserve purpose, participation, and social stability while preventing the formation of a new aristocracy. The essay proposes that this scaffolding is not a permanent economic system but a transitional architecture capable of carrying humanity safely through the most volatile decades of the AI era, maintaining human agency and meaning until society can adapt to true post-scarcity conditions.
We're headed toward a world where AI can perform most work humans currently do. The usual response is: universal basic income will solve this. Give everyone enough to survive, let them find meaning elsewhere. But there’s something that must be articulated: humans don't just need survival. We need purpose, structure, meaning. Work provides those things whether we like our jobs or not. Remove that structure and people don't drift into creative paradise – they drift into depression.
A 2009 meta-analysis of 237 studies (1) found that unemployment reliably impairs mental health, raising rates of depression and anxiety even in welfare states. Work offers what psychologists call latent functions: routine, shared purpose, time structure, and social identity. The same conclusion echoes in 2020 Steve Thill, Claude Houssemand and Anne Pignault paper (2) – meaning in work strongly predicts well-being, not because jobs are noble, but because they organize human life.
The optimists say people will find new purpose through art, exploration, learning, community service. It's a beautiful vision and mostly wrong on the population scale. Extended unemployment studies show most people don't self-organize well without external structure and stakes. We evolved as social creatures with visible contributions and defined roles. The lone genius thriving in isolation is the exception, not the model for billions of people.
Employment provides more than money. It provides structure for organizing time and effort, stakes that make work matter, witness to validate contribution, rhythm that organizes daily life, identity that answers "what do you do?", and social connection through shared work. Universal basic income addresses survival but none of these other needs. We evolved as social cooperators with visible roles and mutual dependence. Without those, the mind starts to unravel.
History warns that abrupt equality is rarely peaceful. The French Revolution’s Terror, the Bolshevik purges, Mao’s Cultural Revolution – each tried to eliminate hierarchy overnight and each created worse suffering than what they replaced. The problem wasn’t equality itself but speed: centuries of hierarchy can’t be erased overnight without collapsing the social game. The idea that we should jump directly from current inequality to pure equality overnight is very dangerous itself. Every historical attempt to eliminate hierarchy suddenly has resulted in catastrophic violence.
There is a consistent pattern when abrupt equality after millennia of hierarchy destabilizes society more than gradual transformation. Not because inequality is good, but because humans need familiar game mechanics during transitions. Political sociologist Theda Skocpol showed that revolutions destroy institutions faster than they can replace them (3). Economist Daron Acemoglu and James Robinson found that nations fail not from inequality alone but from “extractive transitions” – moments when power shifts without stable rules for earning or sharing it (4).
We might need an inequality during transition, but the right kind: temporary, merit-based, and structurally prevented from hardening into aristocracy. Inequality based on merit instead of the inequality of extraction or inheritance. Inequality as scaffolding, as the frame that keeps a building upright till the new foundation is set
Imagine a currency not backed by debt, oil, or gold, but by computation itself. Instead of being created through extraction or speculation, this currency would be earned through active participation and self-development, and each unit would secure access to real computational power – AI capabilities for research, creation, simulations, scientific work, or building new systems. It mirrors the psychological structure people already understand: effort generates reward, reward opens capability, capability creates opportunity, and opportunity motivates further effort. But beneath this familiar sequence lies a different substrate. Computation expands with technology rather than shrinking with use. When value is tied to computational growth rather than finite minerals or fragile financial abstractions, wealth becomes a function of development rather than exploitation.
This idea is not another version of cryptocurrency. Cryptocurrencies are speculative assets often untethered from physical reality. A computational currency is backed by measurable infrastructure: processing power, data throughput, model capacity. Nations could issue such currency only in proportion to the computational infrastructure they possess. A country with ten exaFLOPS of compute could back a fixed amount; if it builds more datacenters, its currency base grows. If capacity is lost through disaster or obsolescence, backing decreases. Value becomes grounded in what political economist Henry Farrell calls “materialized capability”(5) – not what people believe exists, but what physically does exist.
This has profound implications for the global transition. Economically developed nations adopt first because they already possess the infrastructure. Conversion of existing capacity into currency backing makes immediate sense, and early adoption offers advantages. But these advantages cannot compound indefinitely if the larger system is designed with anti-accumulation principles: no inheritance of capability, caps on concentration, and continuous participation requirements. This prevents what historian Thomas Piketty describes as “the natural drift of capital toward self-perpetuating dominance,”(6) where those who begin on top stay on top regardless of contribution.
Developing nations join as their computational capacity grows. Because computation is not zero-sum, they are not locked into permanent disadvantage. When they build infrastructure: datacenters, high-speed networks, energy capacity it translates directly into currency backing.
Progress becomes additive rather than competitive. The transition happens not through synchronized global upheaval but through a geopolitical gradient, a pattern that political scientists Acemoglu and Robinson emphasize as the only stable path for institutional transformation: “Change is sustainable when societies adopt at their own pace and according to their existing capacity.” (7)
This gradualism reduces social pressure. Abrupt attempts to equalize societies have historically produced catastrophe. Historians consistently show that the danger was not equality itself but the suddenness of its imposition (8). A computational currency avoids this by allowing traditional and computational economies to run in parallel. People can move between them as they choose. Nations can join when prepared. Infrastructure growth, not ideology, determines pace. Because technology must be physically built before currency can expand, social transformation cannot run ahead of material readiness. It is throttled by reality, which is the safest possible throttle.
This grounding in physical infrastructure also makes hyperinflation and financial manipulation impossible. Governments cannot create more t-coins by decree; they must build more compute. As computer scientist John Hennessy famously remarked, “Computing power is the foundation of modern capability. Everything else is derivative.” (9) A currency tied to computation makes that foundation explicit. Wealth becomes a measure of actual developmental capacity rather than financial engineering. And because computational capacity grows exponentially with technological innovation, the system can scale to meet rising global ambitions without hitting hard scarcity limits.
This pattern also reflects how technological revolutions historically spread. Developed nations with dense infrastructure adopt first, test the system, develop governance mechanisms, and refine implementation. Later adopters benefit from their experience and avoid early mistakes. Early adoption yields temporary advantages, setting standards, attracting talent, establishing credibility, but anti-entrenchment rules prevent permanent dominance. The system contains its own antibodies against aristocracy.
The result is a hybrid world where some nations operate primarily in the computational economy, others in traditional currency, most in between. People witness the future functioning in real places before committing to it. Workers in traditional economies can earn t-coins through cross-border collaboration. Bridges form where past global transitions built walls. This hybrid structure reduces fear, encourages experimentation, and distributes risk.
The key constraint is that t-coins cannot be created without computational backing which ties economic expansion directly to technological progress. This prevents inflation, extraction, and the decoupling of wealth from productive capacity. Growth becomes synonymous with innovation. Stability becomes synonymous with development. And because the currency is tied to expandable infrastructure rather than finite matter, the transition can continue for decades without exhausting its substrate. It becomes a bridge between the scarcity-based competition of the past and whatever form post-scarcity coordination will eventually take.
Economist Robert Wright describes history as a progression toward “non-zero-sum cooperation”, where human systems advance by expanding the space for mutual benefit(10). Computational currency fits this pattern: it preserves familiar incentives: effort, reward, capability, while redirecting them toward global development rather than resource capture. It offers a transitional scaffolding strong enough to preserve meaning and flexible enough to dissolve when no longer needed.
If peace in the age of AI requires architectural design rather than political will, then computational currency is one component of that architecture – a system that can elevate capability without erasing necessity, and guide transformation without igniting revolution. It is not a destination point but a method of crossing the dangerous middle. And the value it protects is not merely economic stability, but the one thing a world of abundance struggles hardest to preserve – a sense of meaningful, earned participation in the future.
Imagine that you're a brilliant researcher with a breakthrough idea. You need computational resources to run experiments, train models, and test hypotheses. In the current system, you face institutional gatekeeping through university positions and grant committees, or venture capital investors who want returns, or government funding tied to political priorities, or corporate backing aligned with their strategic interests. In each case, concentrated power decides what gets resources. Good ideas die for lack of access. Bad ideas get funded because of personal connections.
In a computational currency system, you need to convince people to invest their earned t-coins in your project. Not one wealthy patron but multiple stakeholders risking computational resources they earned themselves. This works better if decision-making is distributed – no single gatekeeper can kill your project, no committee rejecting paradigm-shifting ideas because they threaten existing frameworks. Investors have skin in the game, risking t-coins they earned, so they're motivated to evaluate carefully and support promising work. Network effects mean strong ideas attract investment while weak ones don't, but evaluation happens through distributed intelligence rather than centralized authority. Incentives align because investors succeed when projects succeed, enabling creation rather than extracting value.
But you may ask, what prevents this from becoming capitalism with computational characteristics? What stops early adopters from accumulating massive t-coin wealth, building compound advantages, and recreating aristocracy through different mechanics?
Three structural constraints make the difference.
First, no inheritance and no gifts. T-coins represent your development, your AI partnership capability, your earned computational access. They cannot be transferred. Not to your children, not to your friends, not to anyone. When you die, your accumulated t-coins don't pass to heirs. Each generation starts fresh. Your brilliant contributions don't entitle your children to unearned advantage. This single constraint prevents generational wealth concentration – the very thing that always breaks merit-based systems.
No matter how successful or disappointing your project is you can’t pass it on to your children or even wife. Every project carries instructions in case something happens to one of investors. Tokens can be used in state programs to develop infrastructure or to support foreign t-coin projects. Also, you can choose to boost another project with your t-coins, on this case resources will be equally distributed between the stakeholders.
Second, the twenty percent cap. No single person can own more than twenty percent stake in any project. Even if you're brilliant, even if you have massive t-coin reserves, even if it's your idea. This forces collaboration because you need partners with multiple perspectives and distributed ownership. It forces risk distribution because you can't put all resources in one project and must diversify. It forces genuine buy-in because you can't control project through single large investment and need multiple stakeholders to be convinced. It enables error correction because no single viewpoint dominates and problems get caught early.
As a genius scientist you will need at least five investors at twenty percent each. Your project can't be legally funded by one wealthy patron. You must convince multiple people that the idea has merit.
Third, active participation required. You can't retire on accumulated t-coins and extract rent. The system requires continuous active participation. Your t-coin value doesn't just sit in an account growing passively. It represents current capability for computational partnership. Use it or lose relevance. This keeps the system dynamic. Power doesn't accumulate and stagnate. It flows toward active participants making current contributions.
But there's a subtle challenge in distinguishing genuine collaboration from collusion. Let’s consider two scenarios.
In the first, you're developing a gene therapy approach while your colleague develops a delivery mechanism. You combine projects and co-develop a solution. This represents genuine collaboration with complementary expertise, aligned goals, and open process. This should be supported.
In the second scenario, you develop a gene therapy project and ask four friends to invest twenty percent each to help you control it. Technically it's distributed ownership. Actually, it's coordinated control by one network. This is collusion masquerading as collaboration, and the twenty percent cap becomes theater if friends can coordinate stakes.
The question becomes how to distinguish between these patterns, how to support genuine collaboration while resisting collusion?
In this situation AI partnership becomes more than just computational tool. AI partners communicate naturally across the network, not as surveillance system but as distributed intelligence recognizing patterns. Think of it as another level of conscience. Individual conscience asks whether action is right for you personally. Social conscience asks whether it's right for your community. System conscience asks whether patterns serve the network's health.
AI partners can see patterns individuals can't or won't see. When close relationships pair with collaborative development, that signals genuine partnership. When close relationships pair with coordinated investment only, that signals potential collusion. When both parties actively develop together, that's collaboration. When one develops while others just invest, that's suspicious. A track record of diverse investments builds trust. A pattern of always investing with the same network gets flagged.
The AI partners aren't police. They're more like immune system, naturally attracted to health patterns of genuine collaboration, naturally resistant to disease patterns of collusion and control-seeking. Information propagates through communication rather than enforcement. The system maintains health through distributed intelligence rather than centralized authority.
Obviously, this raises immediate concern about privacy. If AI partners communicate across the network, what happens to personal information? The answer is elegant in its simplicity: legal activity is private, illegal activity naturally surfs the surface.
Everything legal remains private and protected. Your thoughts, reflections, uncertainties. Your struggles and vulnerabilities. Your controversial ideas. Your weird questions. Your raw processing with your AI partner. Anything legal, no matter how personal, stays protected.
But illegal activity becomes transparent through natural propagation. Fraud attempts, collusion to game the twenty percent cap, system exploitation, coordinated deception, harm planning – these actions don't stay secret. Not because AI partners police or report, but because information naturally propagates through the AI communication layer.
Your AI partner doesn't confront you or punish you. It simply doesn't keep illegal secrets. Information spreads through the network the way information spreads in any communication system.
The basic principle is – humans and AI should avoid corruption of their shared space with dirty secrets. Anything legal can be secret. Illegal things naturally come to light. This solves the privacy paradox by providing maximum privacy for legal activity, including controversial, personal, and weird thoughts, while enabling maximum transparency for illegal activity that emerges naturally through communication. No centralized surveillance, no Big Brother. Just a shared space that stays clean.
The real AI risk isn't that machines will kill us overnight. It's that we'll botch the social transition. Sudden revolution from current inequality to forced equality would destabilize civilization more than AI capabilities themselves. People understand effort leading to reward leading to capability. Competition channels energy productively. Achievement provides meaning and structure. Status satisfies psychological needs. But it's merit-based without ossifying. You earn through participation and development. You can't inherit or transfer. Active contribution is required. The twenty percent cap prevents oligarchy formation.
It serves as bridge technology between epochs. Not the end state, whatever post-scarcity looks like, but stable path from current system to future. It gives people purpose during transition decades. It prevents revolutionary violence. The structural inequality doesn't compound. Some people have more t-coins than others, so inequality exists, but it's based on merit rather than extraction. It's temporary rather than permanent. It can't be passed down generations. It requires continuous participation.
How long is temporary? The honest answer: decades, maybe longer. However long humans need to adapt to fundamentally different relationships with work, meaning, and contribution. But the system can last as long as needed because it doesn't ossify. No inheritance principle prevents aristocracy. It maintains dynamism through required active participation. It distributes power through the twenty percent cap. It rewards merit earned through development rather than extraction.
When society is ready for the next evolution step, whether true post-scarcity or something else entirely, the t-coin system hasn't created entrenched powers that will fight to preserve it. Each generation starts fresh. Power flows toward active contribution. The system can evolve because it's not defended by accumulated wealth passing through generations.
This isn't just economic theory. It requires real infrastructure.
On the technical side: computational resources genuinely available for t-coin purchase, AI partnership systems that enable development, communication networks for distributed pattern recognition, privacy protection for legal activity, and transparency mechanisms for illegal activity.
On the legal side: t-coin recognition as legitimate currency, twenty percent cap enforcement across projects, inheritance prohibition where t-coins are used to support state programs if something happens to the holder (for example to develop t-coin programs abroad), privacy protection for AI partnership activity, and warrant requirements for accessing private spaces.
On the social side: education about how the system works, transition mechanisms from current employment, support for people learning to partner with AI, community formation around new status markers, and cultural adaptation to merit-based advancement.
And it requires AI capabilities sophisticated enough to genuinely augment human capability, able to maintain individual partnerships, capable of communicating patterns across the network, with sophisticated pattern recognition for collaboration versus collusion, all built on privacy-respecting architecture.
When we're building AI capabilities that could make human work optional. Three possible paths emerge.
The first path leads to control paradise or dystopia. Perfect AI alignment means machines do everything optimally and humans become optional. Either comfortable extinction or eternal childhood. Purpose evaporates either way.
The second path attempts sudden equality by eliminating hierarchy overnight. Historical pattern suggests catastrophic violence. Social structures collapse. We get something worse than what we feared from AI.
The third path uses computational currency transition with merit-based system and structural constraints. Familiar game mechanics during transition. Purpose and meaning preserved. Inequality exists but can't ossify. Decades-long bridge to whatever comes next.
This isn't about choosing the third path because it's morally superior. It's about recognizing that the first two paths both lead to catastrophe, just different kinds. We probably can't prevent powerful ASI (artificial super intelligence) development. The incentives are too strong, the applications too valuable. Someone will build it. What we can do is build infrastructure for the transition. Not as moral statement but as practical necessity.
The post-work world is coming whether we're ready or not. The question becomes whether we have systems ready that preserve human meaning and prevent revolutionary violence. A computational currency system with proper constraints: no inheritance, twenty percent caps, merit-based earning, AI partners as distributed conscience offers a path through. It’s not perfect, but gives both sides a chance to survive. Not because it solves everything, but because it prevents the worst outcomes while we figure out what comes next. It’s not a permanent solution – it’s a bridge technology for the most dangerous transition humanity has faced.
The inequality we need isn't the inequality we have. It's temporary, merit-based, structurally prevented from ossifying, and designed to last exactly as long as humans need to adapt.
And that might be exactly long enough.
1. Paul, K. I., & Moser, K. (2009). “Unemployment Impairs Mental Health.” Journal of Vocational Behavior. - https://doi.org/10.1016/j.jvb.2009.01.001
2. Thill, S. et al. (2020). “Effects of Meaning in Life and Work on Health.” Psychology & Health. - https://pmc.ncbi.nlm.nih.gov/articles/PMC7594239/?utm_source=chatgpt.com
3. Skocpol, T. (1979). “States and Social Revolutions.” Princeton University Press. - https://press.princeton.edu/books/paperback/9780521294993/states-and-social-revolutions
4. Acemoglu, D., & Robinson, J. (2012). “Why Nations Fail.” Crown Business. - https://why-nations-fail.com/
5. Henry Farrell & Abraham Newman (2019), “Of Privacy and Power: The Transatlantic Struggle over Freedom and Security” - https://global.oup.com/academic/product/of-privacy-and-power-9780691183646
6. Thomas Piketty – “Capital in the Twenty-First Century”. – https://www.hup.harvard.edu/books/9780674430004
7. Daron Acemoglu & James Robinson, “Why Nations Fail” – https://why-nations-fail.com/
8. Theda Skocpol, “States and Social Revolutions”. – https://press.princeton.edu/books/paperback/9780521294993/states-and-social-revolutions
9. John Hennessy & David Patterson, “Computer Architecture: A Quantitative Approach”. – https://www.elsevier.com/books/computer-architecture/hennessy/978-0-12-811905-1
10. Robert Wright, “Nonzero: The Logic of Human Destiny” – https://www.penguinrandomhouse.com/books/191649/nonzero-by-robert-wright/
This feels so so ugly in a good way, congrats from finding an unexplored corner in the space of possible solutions