Russian roulette: Why you can have a deterministic potential-outcome framework, or an asymmetric utility function, but not both

2024-12-16

AI Summary

This paper addresses a fundamental conflict between asymmetric utility functions—such as the “first, do no harm” principle—and the standard deterministic potential outcomes framework in medical decision-making, which gives rise to logical paradoxes in counterfactual settings (e.g., Russian roulette). The authors rigorously prove that, under asymmetric utility (i.e., differential weighting of harms versus benefits), potential outcomes must be modeled as random variables—not deterministic values. Integrating potential outcomes theory, decision analysis, and formal paradox reasoning, the paper establishes that a stochastic potential outcomes framework is the *only* approach ensuring compatibility between ethical constraints and decision-theoretic consistency. This resolves longstanding logical inconsistencies inherent in asymmetric-utility modeling and fundamentally revises foundational assumptions in medical decision theory. The work provides a rigorous, ethically grounded formal basis for modeling clinical decisions where moral considerations are paramount.

Technology Category

• Reasoning under Uncertainty: Decision/Utility Theory
• Game Theory and Economic Paradigms: Mechanism Design
• Knowledge Representation and Reasoning: Preferences

Application Category

• User Modeling, Personalization and Recommendation: Accountability, Transparency, and Ethics for personalization
• Economics, Online Markets and Human Computation: Fairness and ethical considerations in crowd work and in human-in-the-loop AI systems
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

It has been proposed in medical decision analysis to express the ``first do no harm'' principle as an asymmetric utility function in which the loss from killing a patient would count more than the gain from saving a life. Such a utility depends on unrealized potential outcomes, and we show how this yields a paradoxical decision recommendation in a simple hypothetical example involving games of Russian roulette. The problem is resolved if we allow the potential outcomes to be random variables. This leads us to conclude that, if you are interested in this sort of asymmetric utility function, you need to move to the stochastic potential outcome framework. We discuss the implications of the choice of parameterization in this setting.

Problem

• Resolving paradoxical decisions from asymmetric utility functions
• Modeling stochastic potential outcomes when SUTVA fails
• Addressing counterfactual-dependent utilities in medical decision analysis

Innovation

• Stochastic potential outcomes framework adoption
• Abandoning SUTVA for random variables
• Parameterization choice implications discussion