Living systems reliably achieve stable outcomes despite noise, damage, and incomplete information. Developing tissues regenerate correct anatomy, non-neural networks learn and remember, and evolution itself selects architectures capable of navigating uncertain environments. While these phenomena are well documented experimentally, they remain theoretically fragmented across biology, physics, and cognitive science. This preprint presents Living Information Theory II, an extension and deepening of Living Information Theory (LIT), which formalizes persistence under noise as a governing principle for cognition, morphogenesis, and agency across biological and artificial systems. We introduce the Information Geometric Substrate (IGS) as the necessary mathematical structure implied by LIT: a constrained manifold on which system dynamics, robustness, memory, and control are geometrically encoded. Building on extensive experimental work in developmental bioelectricity—particularly the research program led by Michael Levin—we show that morphogenesis is best understood as goal-directed navigation on IGS, where attractors correspond to viable anatomical states and bioelectric networks act as high-leverage control coordinates that deform basin structure. In this framework, memory is reconstruction rather than storage, goals are dynamical invariants rather than representations, and learning corresponds to flow reconfiguration under noise. The paper integrates: A law-level account of persistence and compression (Living Information Theory), A formal geometric substrate forced by noise-filtered dynamics (IGS), A mechanistic explanation via selection by noise, Empirical grounding in bioelectric control of growth, regeneration, and pattern memory, A unification of morphogenesis, learning, and evolution as the same navigational process operating across timescales. We further connect IGS to measurable causal integration, top-down control, and competency-based definitions of intelligence, yielding testable predictions for developmental biology, regenerative medicine, artificial life, and adaptive AI systems. This work is intended as a foundational, law-level contribution rather than a domain-specific review, and is released as a preprint to enable immediate use, critique, and extension.
T HUNT (Sat,) studied this question.