Artificial Intelligence and the Epistemology of Economic Order

Leave a reply

Subjective Value Theory and the Limits of AI-Driven Production Planning

Abstract

This article examines claims that artificial intelligence (AI) can substantially improve economic coordination and, in some accounts, enable production systems with limited consumer participation. Drawing on the economic calculation debate and the subjective theory of value, the analysis evaluates whether AI can substitute for the informational and allocative functions performed by market exchange. The article contrasts centralized, model-driven approaches to production planning with decentralized coordination through prices, private property, and voluntary exchange. It argues that AI may enhance forecasting, process optimization, and decision support, but it cannot independently resolve the epistemic problem of valuation in the absence of market-generated price signals. The paper concludes that the managerial and policy relevance of AI lies in augmenting organizational decision-making within market institutions rather than replacing the institutional conditions that make economic calculation possible.

Keywords: artificial intelligence, economic calculation, managerial decision-making, production planning, price signals, subjective value theory, market coordination, central planning

Introduction

The rapid diffusion of artificial intelligence (AI) has intensified debate about its implications for productivity, employment, and organizational decision-making. One stream of commentary emphasizes the risks of labor displacement and social dislocation (Islam, 2026), whereas another highlights AI’s potential to increase efficiency and transform productive systems (Diamandis, 2025). A stronger version of the latter claim suggests that AI could eventually manage production with minimal reliance on consumer choice or decentralized market coordination. This article evaluates that proposition by positioning contemporary claims about AI within the long-standing debate on economic calculation. In doing so, it asks a management-relevant question: can AI replace, rather than complement, the informational role of prices, private property, and exchange in guiding production decisions?

Literature Review

The literature on economic coordination has repeatedly examined whether complex production can be directed through centralized design rather than decentralized exchange. Socialist and statist traditions have argued that expert-led planning can correct perceived inefficiencies in market systems, and contemporary advocates increasingly extend this claim to AI-enabled planning tools (Ramana, 2026). More expansive versions of this argument contend that automation will reduce the significance of employment and monetary exchange altogether (Rogelberg, 2026). Against this view, Mises’ critique of socialism remains analytically relevant because it frames coordination not merely as a computational problem but as an institutional problem: without private property and voluntary exchange, there is no reliable basis for economic calculation (Mises, 1920). For management research, this distinction is important because it separates improvements in analytical capability from the institutional mechanisms that generate actionable information for decision-makers.

Analysis

The core limitation of AI-driven production planning is epistemic rather than purely technical. Production decisions require judgments about what should be produced, in what quantities, and with which combinations of scarce resources. These judgments are not fixed optimization problems; they depend on shifting preferences, local knowledge, and opportunity costs that emerge through decentralized interaction. Although AI can process large volumes of data and improve forecasting accuracy, data abundance does not by itself determine which ends ought to be prioritized. From a management perspective, this means that algorithmic systems may improve operational efficiency, but they cannot independently establish the value hierarchy that guides resource allocation across competing uses.

This problem becomes clearer when considered through the subjective theory of value. Earlier objective approaches, including labor- and cost-based theories of value, treated value as if it were inherent in goods. By contrast, the subjective tradition associated with Carl Menger holds that value is assigned by individuals in relation to their preferences and circumstances (Mortell, 2021). Exchange is therefore not a secondary feature of the economy but the mechanism through which these differences in valuation become socially intelligible (Goulart, 2024). For managerial analysis, the implication is that valuation cannot be inferred exclusively from technical inputs, production capacity, or predictive models; it must be understood in relation to choice, exchange, and context.

Discussion

In institutional settings where private property and voluntary exchange are absent, economic calculation becomes fundamentally impaired. If AI were assigned comprehensive control over production, it would operate in the role of a central planner and would lack access to market-generated signals that reflect relative valuations. This restates Mises’ argument in contemporary terms: absent exchange, resource allocation cannot be validated against actual preferences and trade-offs (Mises, 1920). The relevance for management is that even highly capable decision systems remain dependent on the quality of the institutional environment in which they operate. Where exchange is weak or administratively distorted, valuation signals are correspondingly weakened.

Economic calculation compares the expected value of inputs with the realized value of outputs, conventionally expressed through profits and losses (Mises, 1944). These signals do more than record performance; they coordinate decisions across organizations, industries, and time. Prices communicate scarcity, demand, and opportunity cost in ways that no planner, human or artificial, can fully specify in advance. As illustrated by Read’s account of dispersed production knowledge, complex output depends on the coordination of numerous actors whose knowledge is partial and locally situated (Read, 2014). Accordingly, AI is most credible as a complement to market coordination – supporting forecasting, inventory management, and process design – rather than as a substitute for the price system that makes rational allocation possible.

Conclusion

For management scholarship, the central implication is not that AI lacks economic value, but that its value is conditional on the institutional context in which it is deployed. Expanding automation and renewed interest in interventionist policy frameworks (Marr, 2024) make it especially important to distinguish between computational capability and economic coordination. AI can strengthen productivity, analysis, and organizational responsiveness, yet it cannot replace the decentralised processes through which preferences are revealed and resources are economically appraised. Sustainable production systems therefore require not only advanced analytical tools, but also market institutions that preserve exchange, agency, and adaptive price formation.

 

References

  1. Diamandis, P. H. (2025, October 30). AI’s mission: Solving everything. Substack. https://metatrends.substack.com/p/ais-mission-solve-everything
  • Mises, L. von. (1944). Bureaucracy. Yale University Press.
  1. Rogelberg, S. (2026, January 19). Elon Musk says that in 10 to 20 years, work will be optional and money will be irrelevant thanks to AI and robotics. Fortune. https://fortune.com/2026/01/19/when-does-elon-musk-say-work-will-be-optional-and-money-will-be-irrelevant-ai-robotics/

Leave a comment