ORACLE ANALYSIS: Isomorphic Dynamic Programs

TEXT START

We study relationships between dynamic programs by applying conjugacy methods from dynamical systems theory.

THE DISSECTION

This is a paper in pure economic theory doing mathematical archaeology—preserving, refining, and extending the internal coherence of decision-theoretic frameworks that assume an economic agent with meaningful choice. The paper operates entirely within the solved problems of mid-to-late 20th century macroeconomics. Epstein-Zin preferences, Kreps-Porteus, risk-sensitive preferences—these are the furniture of academic macroeconomics circa 1990-2010.

What the paper is really doing: It is proving that structurally equivalent optimization problems yield structurally equivalent solutions, and applying topological/conjugacy machinery from dynamical systems to formalize this. It shows numerical approximation methods improve under isomorphic transformation. This is competent, rigorous work of a particular kind—the kind done when the foundational architecture of the discipline is believed to be sound and the task is to furnish it more precisely.

THE CORE FALLACY (DT FRAMEWORK)

The paper's entire intellectual project rests on smuggled assumptions that are structurally dissolving.

1. The agent exists as an economically relevant entity. The entire framework—dynamic programs, optimality, value functions—presupposes a decision-making agent whose choices determine outcomes. Under P1 (Cognitive Automation Dominance), the agent whose choices drive the program is being systematically displaced by AI systems that solve the optimization problems faster, cheaper, and more accurately. The dynamic program runs; the human running it doesn't.

2. Recursive preferences (Epstein-Zin, Kreps-Porteus) model something that remains. These preference frameworks were designed to handle separation of risk aversion from intertemporal substitution—their intellectual purpose was to make human welfare maximization more tractable and realistic. When AI replaces the human whose preferences are being optimized, the Epstein-Zin aggregator is optimizing a fiction. The value function is still computed. The optimal policy is still derived. The agent it describes is a ghost.

3. Numerical accuracy gains are irrelevant under structural disruption. The paper proudly reports "two orders of magnitude" improvement in value function approximation for a multisector real business cycle model. This is a technical achievement within a paradigm that is itself facing obsolescence. Improving the accuracy of a map of a country that is being redrawn is a fine mathematical exercise; it is not a contribution to understanding the transition.

HIDDEN ASSUMPTIONS

• The economy being modeled has sectors that meaningfully aggregate into a business cycle. (P2: coordination impossibility makes this aggregation increasingly fictional.)
• The agent's preferences are stable and worth modeling recursively. (The paper does handle "time preference shocks"—acknowledging instability—but treats them as perturbations within a stable framework.)
• The optimization problem being solved is the bottleneck in economic performance. (Under DT logic, the bottleneck is productive participation collapse—the optimization problem becomes irrelevant when the majority have no productive role.)
• Mathematical elegance is correlated with economic relevance. (The isomorphic conjugacy framework is elegant. It is also irrelevant to the primary constraint.)

SOCIAL FUNCTION

Prestige signaling and disciplinary self-exoneration.

This is a paper that demonstrates mastery of a sophisticated technical toolkit to peers who value that mastery. It is currency in an academic labor market. Its publication on arXiv in the economics category serves to announce the author's command of advanced macroeconomic theory and dynamical systems methods. The "application" to Epstein-Zin preferences and real business cycle models is the required justification for what is essentially pure mathematical economics. The numerical results are added to satisfy referees who require empirical or computational grounding.

The social function is to keep the academic macroeconomics machinery running—maintaining the intellectual estate, training doctoral students in its tools, producing publications—while the structural conditions the machinery was built to describe are being overridden.

This is not unique to this paper. It is the default state of theoretical economics in the Discontinuity period. The field produces rigorous, technically sophisticated work that is increasingly disconnected from the economic reality described in the DT framework.

VIABILITY SCORECARD (DT Lens)

The paper itself is not obsolescent. The framework it refines is.

VERDICT

Technically airtight. Structurally marginal. Intellectually nostalgic.

The paper proves an elegant result about the mathematical structure of dynamic programs and delivers genuine numerical improvements within its domain. None of this touches the DT mechanism. The paper is addressed to a world where the agent in the dynamic program is the relevant economic actor—and that world is the one being dissolved by AI displacement of productive human labor.

This is autopsy-as-academic-exercise. The patient is still breathing; the paper documents the respiratory mechanics with exquisite precision. What it cannot do is explain why the patient is being unplugged.