Kazutoshi Nagatsu
I am an independent researcher developing Amrita Field Theory (AFT), a unified information-field framework aimed at providing a coherent ontology across consciousness, physics, causality, time, and statistical structure.
AFT models reality as a layered system consisting of:
• 0L (Light / Coherence Domain): global consistency constraints underlying physical law,
• 1L (Information Field): structured informational dynamics associated with identity, meaning, learning, and causal gradients,
• 2L (Phenomenal Domain): spacetime, matter, biological systems, and empirical observations.
Rather than replacing established physical theories, AFT reframes them as effective projections of deeper informational and coherence-level structures. The framework integrates tools from information theory, information geometry, statistical mechanics, and variational principles, and explores their implications for longstanding problems such as:
• the observer and measurement problem,
• the emergence of time and causality,
• non-equilibrium free energy and entropy production,
• agency, learning, and adaptive dynamics,
• the interface between consciousness and physical law.
Recent work (AFT v2.0) consolidates earlier developments into a more formally structured and conceptually conservative core, while keeping speculative or practice-oriented materials clearly separated.
Earlier versions (e.g., AFT v1.5) remain available for historical context, but v2.0 represents the current reference framework.
All papers are released as open-access preprints on Zenodo and are intended as conceptual and mathematical scaffolding for future empirical, computational, and interdisciplinary research rather than as finalized physical theories.
My research interests include:
information geometry, non-equilibrium statistical mechanics, causality, consciousness studies, AI ontology, and unified theoretical frameworks.
AFT models reality as a layered system consisting of:
• 0L (Light / Coherence Domain): global consistency constraints underlying physical law,
• 1L (Information Field): structured informational dynamics associated with identity, meaning, learning, and causal gradients,
• 2L (Phenomenal Domain): spacetime, matter, biological systems, and empirical observations.
Rather than replacing established physical theories, AFT reframes them as effective projections of deeper informational and coherence-level structures. The framework integrates tools from information theory, information geometry, statistical mechanics, and variational principles, and explores their implications for longstanding problems such as:
• the observer and measurement problem,
• the emergence of time and causality,
• non-equilibrium free energy and entropy production,
• agency, learning, and adaptive dynamics,
• the interface between consciousness and physical law.
Recent work (AFT v2.0) consolidates earlier developments into a more formally structured and conceptually conservative core, while keeping speculative or practice-oriented materials clearly separated.
Earlier versions (e.g., AFT v1.5) remain available for historical context, but v2.0 represents the current reference framework.
All papers are released as open-access preprints on Zenodo and are intended as conceptual and mathematical scaffolding for future empirical, computational, and interdisciplinary research rather than as finalized physical theories.
My research interests include:
information geometry, non-equilibrium statistical mechanics, causality, consciousness studies, AI ontology, and unified theoretical frameworks.
less
InterestsView All (8)
Uploads
Papers by Kazutoshi Nagatsu
Contemporary evolutionary theory, especially neo-Darwinism, explains adaptation through random variation and selection, yet leaves persistent explanatory discomfort around repeated convergent solutions and the apparent emergence of highly functional structures.
AFT proposes a three-layer informational ontology—0L (Light domain), 1L (Amrita/information field), 2L (phenomenal biology)—and reframes evolution as a gradient-biased random walk: local mutations can remain stochastic at the organismal level, while the probability of stabilization and retention is subtly biased toward higher field-level coherence, i.e., toward alignment with the 0L coherence attractor.
Within this framework, “inspiration” is treated as a micro-scale evolutionary jump: an individual temporarily synchronizes with a more coherent solution that already exists as a structured possibility in 1L, producing sudden creative or technical breakthroughs.
The paper clarifies how this model avoids naive teleology while still defining a principled notion of evolutionary directionality. It also outlines illustrative biological cases (e.g., neural complexity, sociality, empathy) and proposes observational and experimental routes, including cultural/memetic analyses and tests that manipulate the “meaning-structure” of environments to examine directional bias patterns.