The Realignment

Can capitalism survive? No. I do not think it can.
— Joseph Schumpeter, Capitalism, Socialism and Democracy (1942)

The political axis that organized industrial-era conflict—labor versus capital, regulation versus markets, redistribution versus growth—does not map cleanly onto the Factor Prime transition. The inherited categories will fracture under pressure from a production function that respects none of them.

Consider the constituencies. Libertarians who favor permissionless markets must contend with a transition that concentrates ownership in entities whose positions derive from regulatory moats and network effects rather than competitive merit. Progressives who favor worker protection must contend with a transition where the workers most displaced are the cognitive professionals who constituted their coalition's upper tier. Conservatives who favor tradition and stability must contend with a transition that dissolves the employment relationships and community structures that tradition requires. Environmentalists must contend with a transition that dramatically increases energy demand while potentially enabling efficiency gains that reduce aggregate resource use.

Each position fractures. New coalitions will form along different axes. What Polanyi called the "double movement," market expansion met by protective countermovement, is beginning.(Polanyi 1944) The protective response to the Factor Prime transition has not yet crystallized, but the conditions are assembling.

The hazard curve may not be monotonic. If safety is a luxury good, if richer societies allocate more resources to risk reduction, then the transition exhibits a Kuznets curve: risk rising initially as capabilities outpace institutions, then falling as wealth accumulates and prudent allocation becomes affordable. We may be on the ascending portion now. The distributional fights of the present decade occur at the moment of maximum hazard, when the direction of the curve is still being determined. But this prediction depends on an assumption: that the political system permits the reallocation that theory predicts. Capture dynamics are precisely about the failure of this assumption. Incumbents do not want the safety share to rise; they want the authorization share to rise, with safety as legitimating rhetoric. The Kuznets curve obtains only if capture fails.

The three axes that matter are deployment pace, liability allocation, and surplus distribution.

On deployment pace, the divide separates those who benefit from speed from those who benefit from delay. Frontier capability developers, infrastructure operators, and early-adopting enterprises benefit from rapid deployment; their positions depend on being first to capture automation returns. Incumbent practitioners, professional guilds, and risk-averse regulators benefit from delay; their positions depend on the continued scarcity of what automation makes abundant. The strange bedfellows: the libertarian technologist and the enterprise executive, united in wanting deployment gates lowered, opposed by the union and the professional association, united in wanting them raised. The political conflict here is often misread as safety versus innovation. It is more accurately understood as a fight over the pace at which rents transfer from incumbents to deployers.

On liability allocation, the divide separates principals, platforms, and developers, each favoring a different set of incumbents as traced in the preceding chapter. The political conflict here is often misread as safety versus innovation. It is more accurately a fight among potential defendants over who bears the cost when deployments cause harm.

On surplus distribution, the divide separates those with claims from those without. Equity holders in the actuation layer—datacenter operators, chip fabricators, energy providers, frontier labs—capture returns as labor share declines. Credentialed professionals in authorization-bearing roles capture rents as gatekeepers. Workers in displaced cognitive occupations capture nothing unless policy intervenes.

The policy menu is known. None of the options is obviously correct.

Universal basic income decouples income from employment. Every major pilot to date — Finland's 2017–2018 experiment, GiveDirectly's twelve-year study across 295 Kenyan villages, Stockton's SEED program — found neutral or positive effects on agency, purpose, and well-being; Stockton recipients gained full-time employment at more than twice the rate of the control group.(Hä 2022)(Coltrera 2023) But all were temporary and partial, and recipients knew the program would end. Whether a permanent program at civilizational scale would produce the same result is untested. The theoretical concern — that severing income from contribution may erode the status, identity, and purpose that cognitive professionals derive from work — remains open precisely because the evidence that would close it does not yet exist. The fiscal extraction required to fund UBI at dignified levels will be resisted by precisely the constituencies whose assets would be taxed.

Job guarantee preserves the employment relationship but risks creating make-work that preserves form while evacuating substance. If the guaranteed jobs add no value that anyone would pay for, the guarantee becomes UBI with extra steps and fewer liberties—income disguised as employment, fooling neither the recipients nor anyone else.

Ownership distribution, placing equity claims on computational infrastructure in the hands of those displaced, aligns interests but faces implementation barriers that are not merely political. The assets of the actuation layer (power plants, fabrication facilities, datacenter campuses) are physical monoliths, indivisible in ways that corporate equity is not. Distributing claims on them requires either mandatory dilution of existing owners or public acquisition at scale—neither of which current political economy supports.

Compute or margin taxation targets value capture rather than employment. Its vice is that it creates arbitrage: if taxation is jurisdiction-specific, value capture migrates to lower-tax regimes, and the race to the bottom that characterizes corporate taxation reproduces at the compute layer.

The choice among these options is genuinely indeterminate, but the indeterminacy is structured. The V/C framework specifies the conditions under which each option dominates, and the conditions are empirically distinguishable.

If displacement is concentrated and rapid — if the V/C ordering produces a sharp threshold where large occupational categories automate within a few years rather than a few decades — then UBI dominates, because retraining is too slow and job guarantees cannot absorb the volume. The Finnish and Kenyan evidence supports this for small-scale, temporary shocks; the open question is whether the psychological findings (positive agency effects) survive permanent, economy-wide implementation. The honest bet: if displacement follows the V/C ordering's prediction (high-V/C cognitive tasks automating fastest), the shock will be concentrated in exactly the credentialed professional class least likely to accept a universal cash transfer as adequate compensation for lost status. UBI solves income but not identity.

If displacement is gradual and distributed — if the V/C ordering spreads automation across many domains simultaneously rather than eliminating one at a time — then ownership distribution dominates, because the transition generates surplus that can be shared rather than extracted. The implementation barrier is political, not economic: distributing equity claims on computational infrastructure requires either mandatory dilution or public acquisition, and the political window for either is narrow. The bet: the window opens only during crisis, when concentration has already produced the backlash that makes redistribution politically feasible. Acting before crisis requires the legibility this volume is designed to provide.

If international coordination fails — if the Nash equilibrium is unilateral concentration — then compute taxation is the only option that operates at the jurisdictional level where sovereignty still bites, and its arbitrage problem must be accepted as the cost of any feasible response. The bet: jurisdictional competition for compute capital will produce a race to the bottom that no single country can arrest, and the equilibrium is one or two jurisdictions capturing disproportionate returns while others subsidize access through public infrastructure.

The choice is political. But the conditions that determine which choice is least bad are economic, and the V/C framework makes them legible. The production function determines the pressure; the political system determines the response; the framework determines what the political system can see.

The international coordination problem may be harder than the domestic one. If capital concentration produces growth, and growth produces capacity, the first jurisdiction to defect from a coordinated tax or regulatory regime inherits disproportionate returns. In the limit, a single defector could accumulate enough economic and military capacity to coerce cooperation from any would-be sanctioners. The standard solution to coordination failures, threatening defectors with exclusion, fails when exclusion makes the excluders weaker than the excluded. This is not a failure of will but a failure of the game structure. The Nash equilibrium may be unilateral concentration, regardless of any participant's preferences. The race to attract compute capital may have no stable cooperative solution, only a winner.

Economic analysis addresses income, consumption, and material welfare. It does not address what happens when the activities through which humans derived identity, status, and purpose are performed by systems that experience none of these. The radiologist who spent a decade mastering subtle shadows on a scan, the attorney who built expertise across thousands of cases—each faces not merely income disruption but identity dissolution.

Durkheim's concept of anomie describes the dissolution of shared normative frameworks when traditional social structures erode faster than new ones form.(\'E 1893) The displacement of cognitive labor attacks the occupational identity that served as the primary source of social integration for professional classes. Case and Deaton documented a structurally analogous pattern in their 2020 analysis of "deaths of despair": mortality from suicide, drug overdose, and alcoholic liver disease among white non-Hispanic Americans without college degrees rose steadily from 1999 to 2017.(Deaton 2020) The mechanism was not income loss per se but loss of the social structures—steady employment, community membership, marriage—that provided purpose. If cognitive displacement replicates this pattern for credentialed professionals who have built identity around mastery, expertise, and intellectual contribution, the psychological cost may be qualitatively different but structurally similar.

The empirical evidence on AI-displaced knowledge workers supports the parallel. Sharma et al. (2025) studied 744 workers displaced by AI and found elevated anxiety, depression, and professional identity crisis.(others 2025) Sixty-two percent reported that the loss of professional purpose was more distressing than income reduction. The American Psychological Association's 2023 survey found that 56 percent of workers reported negative mental health impacts from AI-related job changes.(Association 2023) Teutloff et al. (2025) showed a 14 percent decrease in jobs in high-AI-exposure sectors from 2019 to 2024, concentrated among mid-career professionals with 10 to 20 years of domain expertise.(others 2025) The displacement is not random. It targets precisely those who have invested most heavily in occupational identity as the source of meaning.

The displacement of agricultural labor produced generations of dislocation before industrial equilibria formed. The displacement of manufacturing labor hollowed communities organized around production. Whether the displacement of cognitive labor produces liberation or prolonged crisis depends on factors that economic analysis cannot predict: the resilience of social structures, the adaptability of institutions, the human capacity to reconstruct purpose in changed circumstances. This dimension exceeds economic analysis, but intellectual honesty requires acknowledging it. The material consequences are large. The human consequences may be larger.

The default is concentration.

Under multiplicative wealth dynamics—where returns compound on existing holdings—absent redistribution, concentration is not a tendency. It is a theorem.(Peters 2019)

The production function generates returns to scale and network effects that favor incumbents over entrants, early movers over late. The ownership structures that exist at the moment of transition determine who captures surplus over subsequent decades.

The concentration is already visible in the data. Public stock ownership has a Gini coefficient exceeding 0.9; the bottom half of Americans own approximately one percent of equities. Private equity, where AI-native firms concentrate during their growth phase, is more unequal still. The share of corporate capital held by private firms has more than doubled since 2000, rising from eight percent to nearly twenty, driven by intangibles whose value outsiders cannot assess. The shift is structural: twenty-six percent of firms first financed in 1994 went public via IPO; only two to three percent of those first financed in 2000 did.(Farre-Mensa 2020) The AI transition is occurring substantially outside the public markets that democratized access to the gains from the industrial transition. Those without access to private placements, the overwhelming majority, are structurally excluded from the returns.

The Gilded Age provides the template. The positions that mattered in railroads were established in the first two decades: land grants determining routes, terminal facilities controlling urban access, rate agreements allocating traffic. By the time the public understood what had happened, the positions were entrenched. Standard Oil's dominance was not superior refining—competitors could match Rockefeller's processes—but pipeline control and railroad rebates established before regulatory attention arrived. The antitrust remedies took decades; the positions had already generated their returns.

The same dynamic is unfolding now. Hyperscalers have committed to multi-gigawatt nuclear and renewable contracts stretching decades into the future. Fabrication bottlenecks have sold out advanced packaging capacity years ahead, with allocation concentrated among a handful of buyers. Interconnection queues at major grid operators have quadrupled. Each decision appears local: a permitting question, a contract negotiation, a capacity reservation. The cumulative pattern — that these decisions collectively allocate claims on the surplus Factor Prime will generate — is visible only from a vantage point that no single participant occupies. When the pattern becomes legible, the positions will have been established and the costs of reversal will be high. Concentration, once institutionalized, develops constituencies that resist its reversal.

The alternative is not a specific policy but a disposition: scrutiny of the chokepoints where claims are being established, and attention to the distribution of ownership as it forms rather than after it has solidified. Whether this disposition finds political expression depends on organization. Organization depends on legibility. Legibility depends on understanding.

The Factor Prime thesis is an attempt at legibility—a precondition for the organization that alternatives require. It does not prescribe outcomes. It clarifies what is at stake so that choices can be made with awareness rather than by default.

The distribution of surplus is being decided now, by actors who do not know what they are deciding. The default is concentration; alternatives require understanding, organization, and will.