Learning the Value of Value Learning

2025-11-21

AI Summary

Traditional rational choice theory assumes static values, failing to capture how agents’ values evolve through interaction. Method: We propose a dynamic value modeling framework extending the Jeffrey–Bolker utility theory by integrating Bayesian learning and game theory, and introduce— for the first time—the “Information Value Theorem for Value Learning,” formally characterizing the coevolution of values among multi-agent systems. Contribution/Results: We rigorously prove that value coevolution transforms zero-sum games into positive-sum interactions, enabling Pareto-improving Nash equilibria; unify cognitive updating and value revision within a single formal system; and reveal how systematic value adaptation intrinsically enhances societal cooperation efficiency. This work establishes a novel paradigm for AI alignment, multi-agent coordination, and value-sensitive decision-making.

Technology Category

• Humans and AI: Learning Human Values and Preferences
• Game Theory and Economic Paradigms: Cooperative Game Theory
• Multiagent Systems: Mechanism Design

Application Category

• Economics, Online Markets and Human Computation: Social networks and social learning
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Standard decision frameworks addresses uncertainty about facts but assumes fixed values. We extend the Jeffrey-Bolker framework to model refinements in values and prove a value-of-information theorem for axiological refinement. In multi-agent settings, we establish that mutual refinement will characteristically transform zero-sum games into positive-sum interactions and yields Pareto-improving Nash bargains. These results show that a framework of rational choice can be extended to model value refinement and its associated benefits. By unifying epistemic and axiological refinement under a single formalism, we broaden the conceptual foundations of rational choice and illuminate the normative status of ethical deliberation.

Problem

• Extends decision theory to model value refinement under uncertainty
• Transforms zero-sum games into positive-sum interactions through mutual refinement
• Unifies epistemic and axiological refinement within rational choice framework

Innovation

• Extending Jeffrey-Bolker framework for value refinement
• Proving value-of-information theorem for axiological refinement
• Transforming zero-sum games into positive-sum interactions