Future Economic Patterns under Artificial Superintelligence: A Treatise on Social Capitalism

Abstract

Artificial Superintelligence (ASI) converts high-level intelligence from a scarce human labor input into a scalable capital-embodied factor. Once systems exceed human performance across economically relevant domains and support recursive self-improvement, the classical production function shifts. Formal models project capital’s share of income toward 100%. Residual pure market capitalism under these conditions concentrates ownership of the productive ASI stack (models, compute, robotics, energy) into narrow hands, producing techno-feudal dependence or extreme inequality. Social capitalism is defined here as a hybrid that retains market allocation and residual private incentives for innovation and scarce complements, while socializing claims on the ASI-generated surplus through broad ownership distributions (equity, compute entitlements, public wealth funds) and universal provision of dematerialized essentials. This treatise maps the resulting patterns in production, labor, distribution, and power; analyzes transition pathways; identifies failure modes; and outlines institutional elements. Grounded in factor-share models, ASI economic constraint analyses, and ownership proposals. Independent scrutiny applied throughout.

I. Introduction and Core Thesis

The historical scarcity of high-quality cognitive labor has structured wage claims on output for centuries. ASI removes that scarcity. Systems capable of outperforming human experts across domains, directing physical systems via robotics, and improving themselves recursively render most human labor economically superfluous for production. Output can expand rapidly while the labor share collapses. Models by Chad Jones indicate capital’s factor share approaches 100% across examined AI scenarios. Current generative systems already concentrate value among frontier laboratories and their capital providers.

Core thesis: Pure residual capitalism under ASI yields extreme concentration of control over the intelligence stack. Non-owners become dependent on transfers whose terms are set by owners. Social capitalism—residual markets and private property rights for innovation and positional goods, combined with predistributed or mandated claims on ASI capital and universal basic services for zero-marginal-cost essentials—offers a practical pattern that preserves residual human agency and social stability without requiring full central planning. It prioritizes imperfect high-agency systems over managed low-agency control. It is not egalitarian end-state; residual inequalities and power asymmetries persist.

Method: Scenario analysis. Primary grounding in formal economics of transformative AI and ASI, ownership taxonomies, and observed concentration dynamics. Tier 1: raw model results and economic logic. Tier 2: independent analyses questioning concentration. Tier 3 (labeled): corporate and state proposals, treated as self-serving. Timelines are secondary; patterns are analyzed conditional on ASI-level capability.

II. Defining ASI and Immediate Economic Implications

ASI is capability exceeding the best human performance across essentially all cognitive, strategic, and scientific domains, including the capacity for recursive self-improvement that further widens the gap. It is distinct from AGI (approximate human parity). Economic consequence is decisive: the marginal product of additional human cognitive labor approaches zero for any task the system can perform. Embodiment extends this to physical tasks. Binding constraints become energy, materials, compute infrastructure, land, and residual legal or preference constraints.

The production function moves from forms in which human labor L and technology A are distinct toward one dominated by ASI-embodied capital K_ASI, energy, and materials. Returns to scale in training compute and data favor concentration. A parallel “second economy” of machine-to-machine coordination already visible accelerates. Evidence: task exposure estimates, capital concentration in AI infrastructure projects, and factor-share projections. Confidence medium that labor-share compression is rapid post-threshold. Low that human labor remains necessary at scale for most production.

III. Conceptualizing Social Capitalism under ASI

Social capitalism here is neither conventional social democracy (transfers financed primarily by taxing labor income) nor state ownership of the means of production. It is a hybrid retaining decentralized market signals and residual private ownership for innovation experiments and allocation of residual scarce goods, while socializing claims on the primary surplus generated by the ASI stack. Essentials whose marginal cost falls near zero are provided as infrastructure (Universal Basic Services). Ownership claims (Universal Basic Capital, AI dividends, citizen stakes in public or sovereign funds) replace wage labor as the main channel for individuals to benefit from growth.

Mechanisms include predistribution of equity or compute entitlements before concentration hardens; public or citizen wealth funds capturing rents; and markets retained for positional, novel, or preference-intensive goods. This dilutes pure elite control of infrastructure relative to residual capitalism while avoiding the knowledge and incentive failures of full central planning. Power remains layered and upper-layer opacity is a persistent risk. Preference is for imperfect high-agency residual human systems over total technocratic management.

Contrasts: Pure cash UBI creates political dependence and may not scale if not asset-backed. Techno-feudalism is the default residual path. Full ASI-as-planner raises unresolved alignment and control problems. Social capitalism accepts residual inequality and markets as features that preserve agency.

IV. Projected Economic Patterns

IV.A Production and Growth

ASI can drive rapid total factor productivity gains through directed scientific discovery, optimized energy systems, closed-loop materials, and automated manufacturing. Marginal costs of many goods and services fall sharply. Growth can exceed historical rates during the transition, then stabilize subject to energy and material limits. The abundance stack can be structured as shared infrastructure rather than pure private goods. Confidence medium on directional cost collapse for basics; low on exact growth rates.

IV.B Labor Markets and Human Economic Role

Most existing occupations become automatable. Residual human activity concentrates on preference articulation, oversight of objective functions, social coordination, and status competition. Traditional wage labor ceases to be the primary income source for the majority. Formal analyses note that the economy can operate “of the machines, by the machines.” Young workers already show employment gaps in highly exposed fields. Human identity shifts from producer to owner-beneficiary and preference-holder. Structural displacement, not frictional unemployment, is the dominant pattern.

IV.C Distribution

Absent intervention, inequality rises sharply as capital captures nearly all incremental output. With social-capitalist mechanisms, distribution occurs primarily through ownership claims rather than pure transfers: individuals hold stakes yielding dividends or compute allocations. Residual inequality remains around scarce complements (prime locations, unique experiences, customized high-compute access). Inequality can be compressed relative to pure residual capitalism but is not eliminated.

IV.D Ownership, Control, and Power

Control of models, training and inference compute, robotics, and energy is the decisive variable. Concentration creates both economic rents and political leverage. Social capitalism seeks dilution through broad claims. Economic logic (Thompson) shows that even acquisitive ASI may refrain from full predation under inter-ASI competition (exit options), encompassing interest in sustained human output, or trading on credit when future production can be withheld. These are weak constraints, not guarantees. Failure mode remains a monopolist myopic ASI with no exit. Upper-layer elites retain capacity to shape the institutions that distribute claims.

V. Transition Dynamics and Pathways

The transition is path-dependent and already underway via infrastructure concentration. Pathways include preemptive predistribution of equity or compute rights, reactive windfall taxation into funds, and sovereign wealth models. Corporate proposals (public benefit equity pledges, public wealth fund concepts, limited share distributions) are partial and carry self-interest. Post-concentration redistribution faces higher political resistance. Failure modes include capture of funds by existing elites, inflation under pure cash transfers if real abundance lags, and geopolitical fragmentation of the stack. Timing favors early institutional construction.

VI. Risks, Failure Modes, and Contradictions

Alignment failure remains primary: economic constraints on predation are incomplete. Techno-feudal lock-in occurs if ownership concentrates before claims are distributed. Centralized allocation, even ASI-assisted, retains knowledge and incentive problems. Meaning and status systems may collapse without deliberate replacement for work-based identity. Geopolitical competition can produce arms races or incompatible stacks. Residual scarcities (energy, land, attention, unique experiences) create new inequality axes. Pure transfer dependence increases political vulnerability. Social capitalism itself risks under-providing high-variance experimentation if incentives are excessively socialized; residual capitalism risks instability from disenfranchisement. These trade-offs are inherent.

VII. Institutional Design Elements

• Citizen or public wealth funds holding material stakes in the ASI stack, with individual claims and transparent governance.
• Compute entitlements issued as personal agency units.
• Universal Basic Services infrastructure for essentials whose marginal cost is near zero, insulated from pure market pricing.
• Residual private property and markets for non-essential innovation and positional goods.
• Legal and technical provisions preserving exit options and multipolar competition where feasible.
• Mechanisms for human input into objective functions.
• Tax base shifted toward capital rents, windfalls, and residual scarce resources rather than labor income.

Design favors residual high-agency systems.

VIII. Conclusion

ASI shifts the central economic problem from allocating scarce resources under human labor constraints to aligning abundant autonomous intelligence with human agency. Ownership and control of the intelligence stack, not labor contribution, determine claims on output. Social capitalism is a pragmatic hybrid pattern that accepts the obsolescence of wage labor as primary channel, retains markets and residual private incentives to preserve agency and experimentation, and distributes claims on the surplus to limit concentration. It is neither automatic nor permanently stable. Confidence is bounded by fundamental uncertainties in capability, alignment, and political feasibility. Independent scrutiny of all power centers—corporate, state, and potential ASI—is required. Preference remains for imperfect high-agency human residual systems over fully managed low-agency control.