Claims on the Surplus

Can capitalism survive? No. I do not think it can.

Surplus does not distribute itself. By the time an economy can see that a new source of production has arrived, much of the bargaining order around it has already been built: assets purchased, licenses granted, standards written, liabilities assigned, and expectations converted into prices. What later appears as the natural return to a factor is often the accumulated result of these earlier decisions.

Computational abundance will not end that history. It will intensify it. If cognitive output becomes cheaper while fabrication, power, distribution, data, and authority remain scarce, a larger share of the gain can accrue to whoever controls those gates. If those constraints loosen, the surplus may move again. There is no single destination, but there is a recurring political question: who holds a claim when the new capacity begins to pay?

Industrial categories do not map neatly onto the conflict. A worker may own shares in the infrastructure that threatens an occupation. A physician may defend a licensing regime as both patient protection and livelihood. A frontier laboratory may argue for public safety while seeking rules that smaller competitors cannot satisfy. A state may subsidize private concentration in the name of national resilience. Capital and labor remain real positions, but neither names a coherent political camp.

Polanyi's double movement is useful because market expansion and social protection develop together rather than in separate eras.¹ New capacities disembed activity from inherited relations; countermovements attempt to restore security and control. The protection may defend people against ruin. It may also protect the very offices through which older distributions were maintained. A computational countermovement will contain both purposes.

Hirschman's account of commerce as a force expected to tame political passion supplies a second warning.² Institutions often receive support for the vice they are expected to displace. Markets were valued partly because calculation might restrain aristocratic violence, yet commercial society generated powers the original wager did not foresee. Computational systems are now offered as correctives to partial judgment, bureaucratic delay, and human inconsistency. They may reduce those failures and create new forms of dependence at the same time.

The conflict will not be decided by choosing innovation or safety in the abstract. It turns on three separable questions. How quickly should a system acquire permission to act? Where should losses fall when it fails? How should ownership of the resulting productive capacity be distributed? Every coalition will answer them differently, and often inconsistently.

Those who benefit from early deployment will emphasize the costs of delay. Those whose assets depend upon existing scarcity will emphasize hazards of entry. Potential defendants will prefer liability to settle elsewhere in the chain. Owners of computational infrastructure will favor broad access to its outputs without necessarily favoring broad ownership of the infrastructure itself. None of these interests invalidates the argument offered on its behalf. It explains why rules that appear technical decide claims on the surplus.

Income, Ownership, and Standing

The familiar policy menu contains no automatic answer.

Cash transfers can separate subsistence from employment and give people bargaining room during disruption. Evidence from limited and temporary programs bears on labor supply and well-being under those conditions.³ It cannot establish the social effects of a permanent national settlement after occupational structures have changed. Income is indispensable. It is not identical to membership, esteem, or a recognized contribution to a common project.

A job guarantee preserves an employment relation and can direct labor toward public purposes neglected by markets. Its danger is not simply "make-work." The deeper risk is an institution that grants income while withholding agency, assigning tasks mainly to preserve the appearance that support has been earned. A guarantee succeeds only if its offices possess useful purposes and participants can influence how those purposes are pursued.

Broad ownership attacks the problem nearer its source. If computation increases returns to a concentrated stock of infrastructure and intangible capability, wider claims on that stock can allow households to share in the gain. Public funds, pension systems, employee ownership, cooperative platforms, and other vehicles offer different routes. Each raises questions about governance, liquidity, risk, and who exercises control rather than merely receives a distribution.

Taxes on profits, rents, land, energy use, or computational capacity can fund adjustment and public goods. Each base creates different incentives and opportunities for avoidance. A tax upon physical capacity is visible but may discourage useful investment or drive movable loads elsewhere. A tax upon rents is attractive in theory and difficult in practice because rents are rarely labeled in accounts. International coordination can limit arbitrage without abolishing rivalry among jurisdictions.

These policies are complements as often as alternatives. Their appropriate mixture depends upon the pace and distribution of displacement, the ownership of scarce assets, fiscal capacity, mobility of the tax base, and the institutions through which people obtain status and voice. V/C may help anticipate where deployment encounters cheap or costly verification. It cannot tell a polity what its members are owed.

That question reaches beyond income. Durkheim used anomie to describe the disorder that follows when established norms lose their hold before another moral order can form.⁴ Case and Deaton's account of deaths of despair links mortality to a long deterioration in work, family, and community among Americans without college degrees.⁵ Neither history licenses a direct forecast for displaced professionals. Together they warn against treating work as a paycheck with a calendar attached.

Occupations organize time, reputation, apprenticeship, friendship, and the experience of being needed. A radiologist or attorney who loses an exclusive cognitive function may keep an income and still lose an identity built over decades. The response cannot be to preserve every task after its usefulness has gone. It must be to construct forms of standing that do not require scarcity of knowledge as their foundation.

Concentration Is Built

Multiplicative wealth dynamics can produce concentration when gains compound upon unequal starting positions.⁶ Piketty's historical argument shows how sustained returns to wealth above aggregate growth can reinforce inherited inequality.⁷ Neither result makes every future distribution a theorem. Taxes, consumption, shocks, competition, institutions, and political action alter the path. What they establish is that broad diffusion of a technology does not imply broad diffusion of its returns.

The present transition begins from unequal ownership. Much frontier development occurs inside private firms or through assets available mainly to large institutions. Ewens and Farre-Mensa document the long decline in the probability that venture-backed firms enter public markets, one reason households can be excluded from growth that occurs before an initial public offering.⁸ Physical bottlenecks intensify the effect because grid positions, fabrication facilities, and data-center sites are allocated before their future value is fully legible.

Yet concentration is not produced by one mechanism. Scale may lower unit cost. Network effects may reward incumbency. Proprietary data may improve a system. Regulation may require a balance sheet large enough to bear liability. Public subsidies may favor established firms. A license may restrict entry. These mechanisms demand different remedies. Calling all of them monopoly would obscure the work each one does.

Authorization capital deserves particular scrutiny because its scarcity is made partly by law. A credential, permit, insured status, or recognized identity can be a productive asset when it allows action to proceed under accountable conditions. It becomes extractive when the holder can charge for access without preserving the public function that justified exclusivity. The remedy is not necessarily abolition. It may be portability, independent certification, public insurance, transparent queues, contestable standards, or a wider class of answerable principals.

The timing is unforgiving. An interconnection agreement, chip allocation, proprietary workflow, or statutory license can create a position whose returns endure after the decision that established it has disappeared from view. Later redistribution addresses income already realized. Earlier governance asks who receives access, on what terms, and subject to what obligation while the asset is still being formed.

That is the political force of Factor Prime. It does not derive justice from physics. It shows why the material organization of computation and the institutional organization of permission cannot be treated as separate subjects. Electricity can be routed into cognitive work. Models can make some of that work reproducible. Computational systems can help choose investments in the next generation of capacity. None of these facts determines who owns the result.

The claims are being written at the gates. They appear as power contracts, procurement rules, model licenses, data rights, insurance exclusions, professional duties, and terms of access. Read separately, each looks local. Read together, they are the emerging distribution of a new productive order.

The surplus may be large. That is not yet a public achievement. It becomes one only when those who live under the resulting order possess more than cheap cognition—when they possess standing in the institutions that decide where cognition may act and a claim upon what its action makes possible.