Prospect Theory as Active Inference: A Metabolic Account of Risk-Sensitive Decision Making

Prospect theory's characteristic patterns (loss aversion, reference dependence, and nonlinearprobability weighting) have generally been interpreted as cognitive biases, i.e. as evidence ofbounded rationality. This paper proposes a conceptual framework for understanding thesephenomena through the lens of active inference and the free energy principle. We argue thatprospect theory's central features are consistent with computationally efficient solutions to decisionmakingunder uncertainty within the thermodynamic constraints of neural computation. Lossaversion implements adaptive precision-weighting of prediction errors, allocating greatercomputational resources to negative deviations that threaten survival. Reference dependenceimplements efficient predictive coding, transmitting only surprising deviations from expectations.Probability weighting reflects optimal precision allocation across the probability range whenmaintaining full Bayesian representations would exceed metabolic budgets. This framework issupported by converging evidence: neuroimaging studies show unified value coding withasymmetric precision for losses; pharmacological manipulations reveal dissociable neurotransmittersystems for value encoding versus loss sensitivity; and metabolic manipulations including hypoxia,glucose depletion, and circadian mismatch modulate prospect theory parameters in predicteddirections. Developmental evidence shows that children display probability weighting patternsopposite to adults, with gradual transformation through experience pointing at calibration ratherthan to genetic determination. We propose that prospect theory patterns reflect how biologicalsystems navigate uncertainty under fundamental energetic constraints, with implications forunderstanding decision-making architecture and reconceptualizing rationality.