The thermodynamics of prediction

TL;DR

This paper explores the deep connection between a system's ability to predict environmental changes and its thermodynamic efficiency, showing that optimal predictive systems operate with minimal energy dissipation.

Contribution

It establishes a fundamental link between information predictive power and thermodynamic dissipation, applicable to systems far from equilibrium including biomolecular machines.

Findings

Predictive information correlates with thermodynamic efficiency.

Nonpredictive information reflects unnecessary model complexity.

Optimal systems must be predictive to minimize energy dissipation.

Abstract

A system responding to a stochastic driving signal can be interpreted as computing, by means of its dynamics, an implicit model of the environmental variables. The system's state retains information about past environmental fluctuations, and a fraction of this information is predictive of future ones. The remaining nonpredictive information reflects model complexity that does not improve predictive power, and thus represents the ineffectiveness of the model. We expose the fundamental equivalence between this model inefficiency and thermodynamic inefficiency, measured by dissipation. Our results hold arbitrarily far from thermodynamic equilibrium and are applicable to a wide range of systems, including biomolecular machines. They highlight a profound connection between the effective use of information and efficient thermodynamic operation: any system constructed to keep memory about its environment and to operate with maximal energetic efficiency has to be predictive.