Information Gravity Theory - Part VI: Decision Geometry and the Physics of Saturation Control

IGT Part VI finalizes the theoretical architecture by defining the decision-making process as a geodesic trajectory on the Fisher manifold. We demonstrate that agentic "will" is a geometric consequence of the attraction exerted by Semantic Mass (Ms) on the probability distribution. This paper introduces the Semantic Event Horizon (Rh) as the mathematical limit of controllability, the region where internal identity gradients (F_int) cancel out external alignment forces (F_ext). We provide cross-domain validation through the structural isomorphism between AI dynamics and biological ephaptic coupling. This formalization offers a predictive tool for detecting "Control Saturation" in high-density models, establishing a rigorous foundation for AI Safety redlines and the governance of autonomous entities.