The Receding Goal: AI, Development, and Class Divides

The artificial intelligence revolution does not simply divide people into optimists and pessimists. It divides those who have the conditions to use the future from those who have to survive it.

What makes AI so difficult to think about is not only its technical power, but the speed with which it arrives in deeply unequal societies. A tool can promise access, productivity, and augmented creativity; but that promise does not mean the same thing for someone with time, capital, education, and room for error as it does for someone who is indebted, precarious, or exposed to automatable work.

That is why both optimistic and pessimistic narratives about AI contain some truth. That is precisely the problem. Artificial intelligence is not simply salvation, and it is not simply catastrophe. It is a powerful technology entering a profoundly unequal world. For that reason, it is not distributed as a single experience. For some, AI appears as a tool of expansion. For others, as a new form of exposure.

The important question is not only whether someone is optimistic or pessimistic. The question is where they are looking from.

A person with capital, education, a professional network, free time, English fluency, economic stability, and room for error can experience artificial intelligence as a multiplier. They can experiment, learn, automate parts of their work, produce more, create businesses, access knowledge that was once unavailable, and turn technological speed into advantage. For that person, the future looks like a toolbox.

A person who is indebted, precarious, without job stability, short on time, without a safety net, and dependent on work vulnerable to automation may experience the same technology very differently. Not as a tool, but as a threat. Not as expansion, but as pressure. Not as an open future, but as yet another system arriving from above to reorganize their life without asking permission.

The optimistic narrative says artificial intelligence will democratize knowledge. And it might. There is something real in that promise: access to tools, translation, learning, augmented creativity, automation of tedious tasks, new forms of production. But for now, it also seems to be democratizing anxiety with admirable efficiency.

The problem is not only the technology. It is the speed of the technology inside a social system that distributes the capacity to adapt unequally.

Adaptation is not free. It requires time, money, education, rest, connection, equipment, language, stability, a professional network, mental health, and room to make mistakes. Exactly what not everyone has. That is why the phrase “just learn to use AI” sounds reasonable in the abstract and cruel in context. Learning a new tool is not the same when you have protected time and savings as when you are working two jobs, caring for children, paying rent, living paycheck to paycheck, and trying not to silently collapse, like someone updating internal software on 3% battery.

Here, a class divide emerges in the perception of the future. For the upper classes, AI is often a form of leverage: more scale, more efficiency, more investment, more automation, more capacity to turn previous resources into additional power. For professional sectors, AI is ambivalent: it can be assistant, accelerator, and threat all at once. For precarious workers, it often appears not as ChatGPT writing poems, but as scheduling algorithms, productivity surveillance, automated customer service, scoring, invisible dismissal, optimized delivery, remote management, and reduced bargaining power.

Artificial intelligence does not arrive only as “intelligence.” It arrives as infrastructure, property, platform, surveillance, capital, and control. The person who owns the infrastructure experiences it one way. The person measured by it experiences it another.

This difference in perception also occurs on a global scale. For decades, expressions like “developing countries” offered a temporal illusion: some countries were further ahead, others further behind, but everyone was supposedly moving toward the same destination. The phrase was paternalistic, but also reassuring. You have not arrived yet, but you are on your way.

Viewed from this new technological paradigm, that promise becomes more unsettling. The time to catch up with the center was never neutral. It was also the time during which the center kept accumulating capital, infrastructure, technology, intellectual property, data, platforms, and institutional power. The goal did not stand still. While some tried to industrialize, others captured the next phases: finance, software, cloud computing, chips, models, artificial intelligence, computational energy. The problem was not simply arriving late; it was discovering that the race was designed to produce lateness.

Before, we were told certain countries were “developing.” Now the promise sounds more like: you are in the process of updating the system, accepting cookies, learning Python, paying for the premium subscription, and not crying. We were sold the possibility of “catching up,” but no one clarified that the goal was not a fixed place: it was paying permanently to keep accessing the next version of the future.

This is one of the most difficult points to untangle: the digital revolution speaks the language of access, but it often reproduces the structure of dependency. A country can have AI users without having technological sovereignty. It can have platform consumers without owning data centers. It can have technical talent without controlling chips, energy, models, cloud infrastructure, capital, or intellectual property. It can “participate” in the future without capturing the main value of the future.

Every new technological wave arrives with the same promise: this time, everyone will have access. Then one reads the fine print and discovers that access requires chips, cheap energy, English, capital, cloud infrastructure, data, political stability, free time, and a spiritual calm no one included in the package. If Toffler spoke of waves, artificial intelligence is starting to look like a washing machine on spin cycle.

Alvin Toffler used the idea of a “third wave” to describe the transition toward a postindustrial and information-based society. The metaphor still works, but it falls short. What we are living through now does not look like one wave, but a surge of overlapping technological layers: the internet, platforms, smartphones, social networks, big data, cloud computing, automation, generative artificial intelligence, agents, robotics, perhaps AGI. Each new generation of models reopens the question of what counts as human skill, what counts as protected work, and what counts as a possible future.

The old industrial revolution transformed muscles, factories, transportation, and material production. The digital revolution transformed information, communication, and markets. The AI revolution is beginning to touch something even more intimate: language, knowledge, judgment, creativity, diagnosis, planning, translation, memory, administration, and decision-making. It does not automate only physical or repetitive tasks; it begins to automate fragments of what many people understood as their cognitive value.

That is why this wave produces so much confusion. It does not threaten only “manual” jobs, as a certain technocratic fantasy once promised. It also enters offices, universities, law firms, newsrooms, creative agencies, marketing departments, healthcare, education, programming, design, and consulting. Suddenly, the boundary between protected work and vulnerable work becomes less clear. The professional who once felt far from the factory discovers that they too can be broken down into tasks, measured, assisted, accelerated, partially replaced, or turned into the supervisor of systems that do in seconds what once justified years of credentials.

This does not mean that all human work will disappear. That prediction is usually too simple. What is more likely, at least in many areas, is not immediate total replacement, but restructuring: fewer people doing more, workers supervising tools, wages under pressure, tasks disaggregated, professions degraded, productivity captured by companies, and a growing demand to remain updated all the time. The future does not always arrive as a killer robot. Sometimes it arrives as a dashboard, mandatory training, and a “new opportunity for professional growth.” Terrifying, but with friendly branding.

This is where optimists and pessimists misunderstand each other. The optimist looks at the capabilities of the tool. The pessimist looks at the social conditions in which the tool will be deployed. One asks: “What can this technology do?” The other asks: “Who controls it, who pays the cost, and who captures the benefit?”

Both questions are necessary. Without the first, we fall into automatic rejection and lose sight of real possibilities. Without the second, we fall into naivete and confuse technical capability with human progress.

Artificial intelligence can help diagnose diseases, translate languages, personalize education, assist people with disabilities, accelerate scientific discoveries, reduce bureaucratic work, open creative possibilities, and give people access to powerful tools from which they were previously excluded. That is not minor. It should be said without embarrassment. Technological optimism is not always propaganda; sometimes it is the legitimate perception of a tool that really does expand capabilities.

But artificial intelligence can also concentrate wealth, displace workers, intensify surveillance, degrade wages, produce dependency, manipulate information, automate discrimination, extract data, erode privacy, and accelerate the obsolescence of skills before people have real time to adapt. Technological pessimism is not simply nostalgia either; often, it is historical memory. People remember that promises of efficiency rarely guarantee rest for those who work. More often, they guarantee more efficiency for whoever captures the surplus.

The question, then, is not whether AI will be good or bad. That question is too small. The question is: good for whom, under what conditions, with what protections, with what ownership, with what distribution of benefits, with what rights, with what time to adapt, and with what democratic capacity for decision-making?

Because technology does not arrive in a vacuum. It arrives in a world of unaffordable rent, unequal healthcare systems, debt-driven education, borders, monopolies, platforms, debt, precarious jobs, slow institutions, and ecological crisis. Saying “AI will increase productivity” without asking who captures that productivity is like announcing rain in a city where some people have roofs and others do not. Yes, water falls on everyone. No, it does not mean the same thing for everyone.

What produces unease is not only that the world is unjust. That, unfortunately, is not new. What overwhelms us is the speed. In the 1980s, the illusion that there was time could still survive: time to develop, to educate, to industrialize, to modernize, to catch up. Today, technological speed makes that promise feel fragile. The goal does not only move; it updates itself automatically.

Contemporary anxiety is born there: from the collision between technological acceleration and human lives that need time. Time to learn. Time to rest. Time to reorganize institutions. Time to protect workers. Time to legislate. Time to think. Time to understand what just happened before the next model makes the previous conversation feel old.

The future arrives faster, but not necessarily better distributed. It is like express delivery, except some people receive tools and others receive the invoice.

That is why the debate about AI needs less abstract fantasy and more material analysis. It is not enough to ask what the technology will be able to do. We have to ask what kind of society is receiving it. A powerful tool in an unequal system tends to amplify inequalities unless there are institutions capable of distributing its benefits and limiting its harms. Technology can open possibilities, but politics decides whether those possibilities become liberation, concentration, or discipline.

The challenge is not to choose between optimism and pessimism. The challenge is to understand what each position is seeing. Optimism sees capability. Pessimism sees power. Optimism sees a tool. Pessimism sees a structure. Optimism sees the future. Pessimism asks who has the material permission to live it.

A more honest reading would have to hold both things at once: AI may be one of the most extraordinary tools humanity has ever produced, and it may also deepen some of the oldest fractures of modern civilization. It can expand collective intelligence and also perfect systems of extraction. It can democratize access and concentrate control. It can help workers and also make them more replaceable. It can free time and also intensify the demand to produce more.

The contradiction is not only in the technology. It is in us, or more precisely, in the systems we have built to distribute power, time, risk, and benefit.

That is why the AI revolution does not simply divide humanity into optimists and pessimists. It divides those who have the conditions to use the future from those who have to survive it.

That is the plate of spaghetti we have to untangle. Public debate tends to mix everything together: fear of change, technical enthusiasm, corporate interests, labor anxiety, educational promises, geopolitics, science fiction, class resentment, marketing, investment, regulation, creativity, and existential panic. All of it together, with sauce and no fork.

But perhaps the main thread is this: artificial intelligence is not only a technological revolution. It is a test of distribution. It forces us to ask whether a society that already distributes housing, healthcare, time, education, and security badly will be able to distribute well a technology that multiplies cognitive capacities.

If the answer is no, pessimism is right.

If the answer can be built, optimism still has a task.

The real debate is not whether AI will change the world. It is already changing it. The debate is whether that change will be another round of concentration dressed up as progress, or a real opportunity to redistribute capacity, time, and dignity.

And that question cannot be answered by a model. It has to be answered by a society.

 

Technology & National Boundaries: A Civilization Mismatch

 

One of the stranger realizations that emerges from studying Big History and complexity theory is that technological progress and social maturity do not necessarily move at the same speed.

In fact, they often appear to move at dramatically different speeds.

Humanity can map distant galaxies, sequence genomes, and train large language models on significant portions of civilization’s accumulated knowledge. At the same time, it remains perfectly capable of organizing itself around tribal loyalties, centuries-old grievances, status competitions, and disputes whose origins predate the printing press.

This creates a peculiar form of cognitive whiplash.

On one scale, we inhabit a civilization of astonishing sophistication. On another, we remain a species of highly social primates navigating incentives, identities, and narratives that would have been recognizable to our ancestors thousands of years ago.

The contradiction is only apparent. Both realities are true simultaneously.

Scott Page would likely describe this as a consequence of complex adaptive systems operating on multiple timescales. Technologies can evolve rapidly while institutions, cultures, and governance structures adapt much more slowly. New layers of complexity emerge long before older layers disappear.

The result is a civilization where the props often feel futuristic but the setting still looks archaeological.

Bronze Age instincts coexist with medieval identities, industrial institutions, global communication networks, and frontier artificial intelligence. The layers accumulate faster than they are replaced.

This observation becomes especially relevant when discussing AI.

Many current debates assume that the primary challenge is technical: building capable systems, ensuring safety, increasing performance, and managing deployment. Those are important concerns. Yet an equally important question sits beneath them:

What happens when technologies begin operating at a civilizational scale while governance remains organized around nations?

The mismatch is difficult to ignore.

The training data used by advanced AI systems is not American knowledge, Chinese knowledge, or Argentine knowledge. It is the accumulated symbolic residue of civilization itself: languages, books, scientific papers, software repositories, journalism, philosophy, art, documentation, and billions of human interactions flowing across borders.

The resource is transnational.

The disruption is transnational.

The governance remains national.

Which is a bit like discovering a new continent and then insisting the most important question is which municipal office should process the paperwork.

And that would be manageable if nations themselves behaved like mature participants in a coordinated planetary project. Unfortunately, we often seem determined to prove otherwise.

We can build systems that synthesize the knowledge of billions of people, yet we still struggle to cooperate across borders, parties, regions, and identities. Not because the problems are always impossibly complex, but because incentives, prestige, short-term interests, and the occasional outbreak of political chiquitaje remain remarkably durable features of human affairs.

There is something profoundly puzzling about it.

A species capable of contemplating the origins of the universe can still become hopelessly divided over symbolic disputes, procedural squabbles, and status contests that, viewed from sufficient distance, look suspiciously small. We no longer argue about the exact same goats that wandered into the neighboring field centuries ago, but we continue to manufacture functional equivalents with impressive creativity and enthusiasm.

Meanwhile, greed has not exactly retired from public life. New technologies arrive, new fortunes emerge, and many leaders discover once again that thinking in terms of the next election cycle, the next quarterly report, or the next personal advantage feels more natural than thinking at the scale of civilization. Not always. But often enough to matter.

The challenge is not that humanity lacks intelligence.

The challenge is that intelligence scales faster than wisdom, and capability scales faster than coordination.

Politicians naturally propose national solutions because nations are where political power resides. Taxation, regulation, ownership structures, and redistribution mechanisms all operate through existing states. Senator Bernie Sanders’ proposal to tax extraordinary AI-driven gains and return a portion of the benefits to the public deserves to be taken seriously in this context. It recognizes something many observers across the political spectrum are beginning to notice: AI systems derive value not only from private investment but also from a vast reservoir of collective human knowledge.

That insight is laudable.

It may even point toward a reasonable path for ensuring that the benefits of increasingly capable systems are shared more broadly rather than concentrated narrowly.

But here comes my “but.”

Even if Sanders’ proposal were implemented perfectly, it would still confront the deeper challenge that the systems themselves operate across borders while the mechanisms for redistribution remain tied to individual nations. A national dividend may help address national consequences. It does not fully answer the civilizational question.

This creates a peculiar asymmetry.

A sufficiently powerful AI system may affect labor markets in dozens of countries simultaneously. It may be trained on knowledge generated by people across the globe. The servers may sit in one jurisdiction, the investors in another, the users in hundreds more. The benefits and disruptions spread through a planetary informational network largely indifferent to political borders.

A similar mismatch appears in public health. We often discuss outbreaks in distant countries as though Marco Polo had just arrived in Venice with alarming tales from a land beyond the edge of the known world. The fact that a pathogen can now cross continents faster than Marco Polo crossed a village somehow does little to diminish that feeling. We continue to treat many global health threats as though they were unfolding on Uranus rather than within the same densely connected civilization we inhabit.

The atmosphere does not care where a molecule originated. Viruses do not carry passports. Increasingly, informational systems appear equally indifferent to national borders.

This does not mean nation-states become irrelevant. Governments still regulate, tax, negotiate, and enforce. Companies remain subject to laws. Infrastructure exists in physical places. Reality eventually cashes out into jurisdictions.

But the scale mismatch remains.

The problem is civilizational.

The available tools are largely national.

Even if every country implemented excellent policies tomorrow, the deeper question would remain unresolved.

Who owns the products of collective learning?

That question is far stranger than it first appears.

AI systems are built using private capital, private engineering, and private risk-taking. Yet they are also built upon public research, open-source software, scientific knowledge, language itself, and centuries of accumulated human culture.

The training corpus looks suspiciously like a civilization-scale commons.

This is why arguments about ownership feel different in the AI era than they did in previous technological revolutions. The debate is no longer only economic. It is epistemic.

Who owns the systems that increasingly mediate knowledge, interpretation, memory, explanation, and attention?

That question begins to sound less like a debate about factories and more like a debate about libraries, universities, communication networks, and the informational infrastructure through which societies think.

Unfortunately, history offers little reassurance that extraordinary capability automatically produces wise outcomes.

A civilization can become extraordinarily capable while using both humans and machines in surprisingly stupid ways.

The Roman world produced remarkable engineering while remaining trapped in recurring political dysfunction. The Industrial Revolution transformed productivity while tolerating extraordinary human misery. The internet connected billions of people and then devoted a meaningful portion of its capacity to outrage optimization.

There is no law stating that intelligence, capability, and wisdom must increase together.

Indeed, they often do not.

The future may not resemble the clean technological trajectories imagined by either utopians or doomers. It may instead resemble a civilization becoming progressively more capable while struggling to coordinate around the consequences of its own success.

A civilization that can train frontier AI systems while remaining politically fragmented.

A civilization that can model climate systems while arguing about basic facts.

A civilization capable of mapping exoplanets while still becoming trapped inside local incentive structures.

And perhaps, if we are being honest, a civilization capable of generating endless new disagreements even after solving some of the old ones. If ancient cities could spend generations arguing over whose goat wandered into whose field, modern societies can certainly invent equally passionate disputes over algorithms, data rights, and digital borders. The names change. The coordination challenge remains.

This is not necessarily a sign of failure.

It may simply be the normal condition of complex adaptive systems.

The truly remarkable fact is not that humans remain tribal, emotional, and imperfect. The remarkable fact is that they have managed to build global systems of cooperation despite those limitations.

Perhaps that is the real lesson of collective learning.

Humanity was never required to become wise before becoming powerful.

It only had to become coordinated enough.

Whether wisdom eventually catches up remains an open question.

The Cerberus Market

 

Commodity, Broker, Consumer: Marx, Keynes, and Smith on AI Capitalism

The economic problem is simple enough to state plainly: if capitalism weakens the consumer, who is left to buy? AI capitalism promises cheaper production, more automation, and more productivity. But capitalism does not run on production alone. It runs on production that can be sold. Someone must have money, freedom, and reason to buy what the system produces.

That is where the contradiction starts. A company can cut labor costs and improve its margins. But wages are also demand. If many companies automate work, weaken bargaining power, and concentrate income, the system may become better at producing and worse at selling. It becomes a beautiful machine with a shrinking customer base.

The same problem appears in platform and AI markets. People are not only buyers. They are also data sources, training material, behavioral signals, unpaid evaluators, and dependent users. The market is not merely selling to them. It is built through them.

The system wants people cheap as workers, rich as consumers, transparent as data sources, dependent as users, and creative as training material. Those demands cannot all be satisfied forever.

The Role Confusion

There is an inherited absurdity in being commodity, broker, and consumer at once, because those roles are supposed to be structurally separate. A commodity is sold. A broker mediates the sale. A consumer buys.

Cerberus works because the three heads share one body. Commodity, broker, and consumer are supposed to be separate market roles because they have different interests. In AI capitalism, they are fused into one subject. The result is not clever integration but structural impracticality: one body is asked to be the value extracted, the mechanism of circulation, and the buyer charged for access.

You are the commodity because your behavior, attention, language, preferences, social graph, and future likelihoods are packaged as value.

You are the broker because your clicks, prompts, shares, corrections, ratings, posts, and interactions help route, train, validate, and refine the system. You are not merely being sold; you are helping organize the conditions of the sale.

You are the consumer because you pay for access, products, subscriptions, recommendations, visibility, productivity tools, identity services, and sometimes even privacy from the same systems extracting from you.

This is more than unfairness. It creates economic confusion. If the person is input, market signal, buyer, and disposable cost all at once, the system has trouble knowing what the person is for. It wants to extract from the person and sell to the person at the same time. That can work for a while. It cannot work cleanly forever.

Marx: The Contradiction Inside Capital

Marx helps because he understood capitalism as a system that creates contradictions from within. Capital wants to reduce labor costs, increase productivity, expand markets, and accumulate profit. But labor is not only a cost. Workers are also consumers, social beings, and the human base through which production is reproduced.

This is the contradiction AI sharpens. Capital wants labor minimized at the point of production and maximized at the point of consumption. It wants fewer workers to pay, but enough consumers to buy. Each firm may rationally automate and cut costs. But if many firms do it at scale, the wage base erodes. The individual capitalist behaves rationally; the system becomes collectively irrational. It is the old contradiction wearing better software.

Marx would also notice enclosure. Shared human knowledge, language, code, art, behavior, and social intelligence become raw material for privately owned systems. The collective output of human culture is turned into proprietary capability. Then that capability is sold back as access. This is not land enclosure in the old form, but it has the same structure: a commons becomes private revenue.

The alienation also mutates. In industrial capitalism, the worker is separated from the product of labor. In AI capitalism, people are separated from patterns of their own lives, expressions, and intelligence, which return as proprietary services, rankings, recommendations, scores, and tools.

Keynes: The Demand Problem

Keynes would ask the blunt question: who has the money to buy what the economy can produce? If productivity rises while purchasing power concentrates, the economy can produce more than ordinary people can afford to consume. That is not abundance. It is imbalance.

The rich do not consume in the same proportion as ordinary households. A dollar shifted from wages to profits does not automatically return as broad demand. It may become savings, asset speculation, share buybacks, monopoly expansion, or investment in further labor displacement.

This is the bakery problem: a bakery that can make infinite bread in a town where everybody is celiac is technically impressive and economically useless. The issue is not whether the bakery is productive. The issue is whether its output can be absorbed.

A Keynesian rescue would require political management of AI productivity gains: redistribution, public investment, shorter working hours, income supports, stronger automatic stabilizers, and institutions that keep productivity gains from concentrating entirely at the top. The technical question is demand. The social question is whether automation becomes shared freedom or private rent.

Adam Smith: The Moral Conditions of Markets

Adam Smith can be rescued, but only if we rescue the real Smith, not the cartoon version. Smith was not simply saying greed magically saves society. His economics sits beside a moral theory of sympathy, justice, prudence, trust, and social judgment. Markets require more than self-interest. They require conditions under which exchange is not domination dressed as choice.

Smith was suspicious of monopolies, collusion, rent-seeking, and merchants who capture public policy for private advantage. He understood that business interests often prefer restriction over open competition. He did not think concentrated commercial power automatically serves the public good.

From a Smithian perspective, platform and AI capitalism are suspect because they distort the conditions of free exchange. A market is not truly free when users cannot understand the bargain, avoid the infrastructure, inspect how visibility is priced, contest data extraction, or negotiate with the systems that mediate their work and social life.

This is where the moral dimension matters. Not Victorian respectability, exactly. Smith belongs to the Scottish Enlightenment, shaped by a Protestant moral world in which sympathy, restraint, justice, and social judgment still mattered. A market with the handshake removed and the fine print promoted to king is not a purified market. It is a predatory one.

Remove Smith’s moral compass from Smith’s economics, and the market becomes a logistics system with no conscience. The mistake is not returning to Adam Smith; the mistake is returning to a mutilated Smith, a Smith stripped of sympathy, justice, and suspicion of commercial power.

The market has something of the old maritime trade route in it: cargo, brokers, ledgers, risk, ports, insurance, and respectable distance from harm. The point is not to flatten historical differences, but to notice the recurring form: human life converted into transferable value, moved through an infrastructure of intermediaries, and morally laundered as commerce. In that register, the person is cargo, navigator, and passenger at once: helping steer the ship, paying for the voyage, and still getting marched onto the plank when margins demand it.

The Disappearing Economic Agent

Modern economics often begins with the rational economic agent, but this premise depends on social conditions the model usually treats as background: trust, information, autonomy, stable institutions, enforceable contracts, and meaningful alternatives.

If capitalism corrodes those conditions, the agent at the center of economic theory disappears. What remains is not a free chooser but a managed subject inside private and public infrastructures. At that point, even production is no longer guaranteed, because production itself depends on coordination, skill, trust, demand, and social reproduction.

Smith’s moral dimension is not decorative. It is part of the market’s operating system. Without it, the rational agent disappears; exchange degrades; demand weakens; productivity loses meaning; and capital becomes control over decaying assets.

When Productivity Loses Its Market

The productivity problem is not only that productivity may fall. The deeper issue is that productivity can lose its ordinary capitalist meaning. In capitalism, productivity matters because more output can become more value. But that only works if output can be sold. Without demand, productivity becomes capacity without realization.

Productivity without demand is a factory on an island, getting more efficient at producing goods no ship comes to collect. The machines may be excellent. The output may be enormous. But the market circuit is broken.

Here productivity needs to be understood in its oldest and most basic sense: the capacity to produce more output with less labor, time, land, energy, or material. That meaning has been with us since the agricultural revolution. But under capitalism, productivity must also pass through the market. It becomes economically meaningful not only when more can be produced, but when that output can be sold, financed, or otherwise absorbed as value.

This is the Hegelian shape of the problem, later sharpened by Marx: the contradiction is not external to the system. It grows from inside it. The same logic that pushes capital to automate labor, weaken wages, and concentrate ownership also weakens the consumer base that makes productivity profitable. Put less politely: even in Gucci shoes, shooting yourself in the foot still hurts.

If the mass consumer weakens, the old civilizational meaning of productivity does not disappear. But its ordinary capitalist channel breaks. Producing more with less is still technically powerful; it is just no longer enough to sustain a consumer market. Capital then looks for projects large enough to absorb capacity and justify investment: defense, energy infrastructure, climate adaptation, data centers, compute expansion, logistics, resource control, administrative automation, elite health, or other megaprojects. Space colonization is the cartoon endpoint of this logic; the nearer versions wear hard hats, uniforms, lab coats, and procurement badges.

This changes the question. The market no longer asks only, who buys the product? It asks, what project can absorb capital, machinery, labor, and legitimacy? When the checkout line disappears, capital starts looking for a construction site.

That is why this is not ordinary consumer capitalism. Productivity becomes less consumer-facing and more project-facing. It serves states, corporations, infrastructure owners, security systems, and elite markets. The public may still be involved, but less as a strong consumer and more as a managed population inside the project.

Three Diagnoses, One Crisis

Marx, Keynes, and Smith point to different parts of the same crisis. Marx says the system undermines its own social base. Keynes says it threatens effective demand. Smith says it corrupts the moral and competitive conditions that make markets legitimate.

Put together, the diagnosis is sharp: AI capitalism may produce too efficiently for a society whose income, autonomy, and moral foundations it has eroded. The problem is not that the system cannot produce enough. The problem is that it may damage the people, institutions, and markets that make production meaningful.

Who Will Buy?

The likely answer is stratification. Wealthy individuals buy premium agency: better AI, better health, better education, better privacy, better security, better lawyers, and better insulation from the systems others must inhabit. Firms buy automation to reduce labor dependence. States buy AI for administration, surveillance, defense, welfare management, policing, and public service automation. Ordinary people receive cheaper, degraded, subsidized, ad-supported, behavior-extractive versions.

So the market may not disappear. It may mutate. The old mass consumer becomes less central. Corporations, states, and wealthy households become the most solvent consumers. Everyone else becomes a managed user base: economically weaker, behaviorally legible, technologically dependent, and still valuable as data, attention, compliance, and political population.

The mall does not vanish; it becomes a members-only logistics hub with a public waiting room. That is the drift from consumer capitalism toward rentier-control capitalism. The system earns less by selling abundant goods to a broadly prosperous public and more by charging access, controlling infrastructure, extracting data, licensing intelligence, managing risk, and selling tools of optimization to those who can pay.

If there is any Smithian hope here, it is not that markets fix themselves. It is that markets can be made legitimate, and kept from becoming self-defeating, only when they are held inside moral and institutional limits: fair competition, public goods, real alternatives, restraints on monopoly, and a social world in which people can still act as agents rather than managed inputs.

Smith does not rescue the system by blessing self-interest. He rescues the question by reminding us that commerce without moral conditions is not freedom; it is organized dependency.

The consumer problem is where Marx's contradiction, Keynes's demand failure, and Smith's moral test meet. Not a pleasant room, but a very clear one.