Agentic Information Theory: Ergodicity and Intrinsic Semantics of Information Processes

TL;DR

This paper develops an information theory framework for understanding how memoryful agents interpret and adapt to complex environments over time, focusing on ergodicity and intrinsic semantics of their information processes.

Contribution

It introduces a novel approach to analyze the temporal information processes of cognitive agents, emphasizing ergodicity and semantics in structured stochastic environments.

Findings

Established basic results on ergodicity of information processes.

Analyzed the semantics of time series of Shannon information measures.

Provided insights into how agents interpret environmental uncertainty.

Abstract

We develop information theory for the temporal behavior of memoryful agents moving through complex -- structured, stochastic -- environments. We introduce and explore information processes -- stochastic processes produced by cognitive agents in real-time as they interact with and interpret incoming stimuli. We provide basic results on the ergodicity and semantics of the resulting time series of Shannon information measures that monitor an agent's adapting view of uncertainty and structural correlation in its environment.