AI-mediated markets introduce computational admissibility as a prerequisite for allocation. In such markets, assets may be excluded from consideration not due to intrinsic value, but due to the inability of computational systems to interpret, verify, compare, or act upon their representations. Representation Capital has previously been defined conceptually as the accumulated stock of machine-readable qualities that may increase the probability of computational admissibility in AI-mediated allocation systems. This paper proposes a formal measurement theory for Representation Capital. It distinguishes between latent Representation Capital and measurable proxies constructed from observable primitives. The framework defines a six-dimensional primitive vector—completeness, accuracy, verifiability, freshness, portability, and actionability—and develops additive, multiplicative, and threshold-augmented indices. The paper distinguishes between soft probabilistic admissibility and hard threshold-based exclusion, introduces a computational admissibility function, derives Representation Yield and Allocation Influence, and presents formal results on admissibility monotonicity, representation bottlenecks, threshold exclusion, and saturation dynamics. The paper does not claim empirical validation. It provides a theoretical measurement instrument and generates testable predictions for future empirical research in AI-mediated markets, computational economics, and agent-readable asset systems.
Marco Patrone (Wed,) studied this question.