When Intelligence Becomes Abundant, Power Becomes Scarce

4 min read

A person walks through a dimly lit automated factory, past robotic arms and self-driving carts, with snow-capped mountains outside the windows at dusk.

Two years ago, I was doing my PhD, training models to solve grade-school math problems. Today, a mathematician has claimed to have proved the Riemann Hypothesis with Claude. Whether the proof holds up remains to be seen, but the pace of progress is surreal. What increasingly occupies my mind is not how intelligent AI will become, but whether our political and economic institutions are prepared for a world in which human intelligence is no longer economically scarce.

Modern capitalism rests on an implicit bargain: most people, lacking ownership of productive capital, secure their share of society’s output by selling their labor. Historically, technological revolutions have preserved this arrangement by creating new forms of employment even as they displaced old ones. But what if AI fundamentally breaks that pattern? If machines eventually outperform humans in nearly all cognitive and physical tasks, the problem is not merely technological unemployment or insufficient consumer demand. It is a crisis of distributive legitimacy. If wealth can be produced with almost no human labor, on what moral grounds should access to that wealth remain conditional on employment? And if production increasingly depends on capital rather than labor, why should ownership of that capital determine who gets to prosper?

This is where Marx’s critique of capitalist property relations becomes newly relevant. The contradiction is not simply that machines might replace workers, but that increasingly collective and technologically mediated production can coexist with highly concentrated private ownership of its benefits. Yet redistribution alone may be insufficient. Rawls’s distinction between welfare-state capitalism and property-owning democracy suggests that guaranteeing a minimum income is not equivalent to distributing economic power. Even a generous universal basic income could leave citizens materially secure but structurally dependent on a small class controlling the productive infrastructure. From the perspective of republican political philosophy, freedom is not merely the absence of deprivation but also the absence of domination. A society in which a handful of corporations control the intelligence, compute, and infrastructure on which everyone depends might be extraordinarily wealthy without being genuinely free.

What makes this especially unsettling is the implication for individual behavior. For most of modern history, education and specialized knowledge have provided a relatively reliable path to economic mobility. Human capital, in the economist’s sense, could be accumulated and exchanged for income, status, and influence. But if intelligence itself becomes reproducible at negligible marginal cost, the scarcity value of many forms of expertise may erode. Knowing how to build something may matter less than owning the infrastructure through which it is built, controlling its distribution, or having the institutional authority to determine how it is deployed. The binding constraint shifts from knowledge to power: access to capital, ownership of scarce assets, control over distribution, political influence, and the ability to coordinate other people. Technical understanding remains valuable, but increasingly as a means of exercising judgment and directing resources rather than as a scarce commodity that can be sold on its own.

From an individual’s perspective, the rational response in the short term may therefore be uncomfortable but straightforward: accumulate as much economic and institutional power as possible while human agency still commands a premium. Build ownership rather than merely credentials, acquire assets rather than relying exclusively on wages, cultivate relationships and influence rather than assuming technical competence alone will remain sufficient. In a world where intelligence is abundant but ownership remains scarce, the returns to capital and control may increasingly exceed the returns to expertise. Yet this creates a profound collective-action problem. What is rational for each individual under the existing rules may be disastrous for society as a whole. If everyone responds to the threat of technological displacement by competing for ownership and power, we may accelerate precisely the concentration that makes the transition politically dangerous. The individual incentive is to secure a place among those who own the machines; the collective imperative is to ensure that a decent human life does not depend on being among them.

Perhaps this is the deeper contradiction of the coming age: technological progress may make economic abundance possible while simultaneously making access to that abundance more politically contested. The problem is no longer simply how to produce enough, but how to justify the allocation of a productive capacity that increasingly operates independently of human contribution. Can ownership of AI-driven capital be democratized without undermining innovation? Can universal basic income provide genuine independence, or does meaningful freedom require a broader distribution of productive assets and political authority? And if intelligence ceases to be a source of scarcity, what becomes of a meritocratic ideal that has long justified inequality through differences in individual ability and achievement? We may discover that what appeared to be rewards for personal merit were always partly determined by the scarcity of particular capabilities, and that technological progress can eliminate that scarcity without eliminating the institutions that distribute its rewards.

I don’t know whether the destination is socialism, some form of property-owning democracy, a reimagined capitalism, or a technological oligarchy. But there is something deeply ironic about a technology that could liberate humanity from the necessity of labor instead compelling individuals to compete even more aggressively for capital and power. Personally, I feel the pressure to play that game while the rules still allow it. Philosophically, I find it difficult to believe that a civilization capable of making human labor optional should organize human dignity around who managed to accumulate enough leverage before it became obsolete. The defining political struggle of this century may not be between humans and machines, but between different groups of humans over who gets to own, govern, and benefit from the machines. And perhaps the hardest question is whether we can build institutions that make individual survival independent of winning that struggle.