Markets for Models

2025-03-04

AI Summary

This paper investigates market mechanisms for predictive models as tradable goods: how firms train, price, and strategically select models (e.g., by introducing bias); how consumers aggregate multiple models to improve predictive performance; and how market structure is non-monotonically shaped by the bias–variance trade-off. Employing statistical learning theory—specifically the bias–variance decomposition—combined with game-theoretic modeling and equilibrium analysis, we show that firms may strategically adopt excessively biased models to deter competition or extract monopoly rents, resulting in systemic efficiency losses. Our analysis uncovers, for the first time, an endogenous “bias-as-barrier” mechanism in model markets, characterizing the structural origins of such inefficiencies. These findings provide a theoretical foundation for antitrust regulation and incentive-compatible mechanism design in AI model markets.

Technology Category

• Game Theory and Economic Paradigms: Mechanism Design
• Multiagent Systems: Mechanism Design
• Machine Learning: Ethics, Bias, and Fairness

Application Category

• Economics, Online Markets and Human Computation: Uses of LLMs and GenAI for marketplace design, bidding, and strategic interactions
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Motivated by the prevalence of prediction problems in the economy, we study markets in which firms sell models to a consumer to help improve their prediction. Firms decide whether to enter, choose models to train on their data, and set prices. The consumer can purchase multiple models and use a weighted average of the models bought. Market outcomes can be expressed in terms of the bias-variance decompositions of the models that firms sell. We show that market structure can depend in subtle and nonmonotonic ways on the statistical properties of available models. Moreover, firms may choose inefficiently biased models to deter entry by competitors or to obtain larger profits.

Problem

• Study markets where firms sell predictive models to consumers.
• Analyze how firms choose models, set prices, and deter competition.
• Examine market outcomes based on model bias-variance trade-offs.

Innovation

• Firms sell predictive models to consumers.
• Consumers use weighted averages of purchased models.
• Market structure depends on models' statistical properties.