Lee Sharks, Johannes Sigil, Rex Fraction
A theoretical work classifying semantic compression into three types—lossy, predatory, and witness—and explaining them within a thermodynamic framework.
Existing semantic economy and semantic physics frameworks describe the system but fail to explain how a single compression event produces field-scale effects. A transfer law is needed to account for the impact of the compression's fuel-ledger structure on shared semantic capital.
Define all semantic operations as compression operations, and distinguish three regimes (lossy, predatory, witness) based on the fuel consumed by compression and the location of unrecoverable cost. Theoretically develop the thermodynamic characteristics of each regime and their connection to semantic economy and physics.
Proposes a classification system for semantic compression and a transfer law that integrates existing frameworks. However, it remains at a conceptual level without empirical validation or concrete applications, thus offering limited contribution from an AI research perspective.