Essays
Standalone essays I have written from my research program
Selected chapters of the Concept Record, rewritten by me so they can be read independently.
Published
What AI Cannot Give Back — 'Satoshi,' the One-to-Billions Fate, and a Nine-Dimensional Bonnō Correction
A structural diagnosis of the unease I have carried after a year of working with frontier AI, and the single-line definition that holds my research program together: not "give the AI something," but "give back to the human what the AI's one-to-billions averaging has stripped away." Positioned as a natural extension of Anthropic's Mechanistic Interpretability.
Buddhism as a Coordinate System — A New Computational Foundation for Understanding the Human
Starting from my methodological concern about neuroscience-centered approaches, this essay lays out my argument for adopting the 2,500 years of internal Buddhist observation as a computable coordinate system. The fundamental purpose of my research program. Level 1 (methodological tier).
The External Prefrontal Cortex (ExPFC) — An AI Device that Externally Substitutes for a Compromised Prefrontal Cortex
Scam victimization redefined neuroscientifically as a temporary functional failure of the prefrontal cortex, and the AI device — ExPFC — that externally substitutes for a compromised PFC. A proposal from applied neuroscience standing at the intersection of neuroethics and AI Safety.
Bonnō Embedding and the Scam-Detection Mechanism — Mapping Text into a Vector Space of Psychological Vulnerability
A detection mechanism built around a function f mapping text into a 108-dimensional vector of mental-affliction stimulation, composite-judged with a law-grounded check on source attribution. Addressed equally to Affective Computing researchers and to anti-scam practitioners.
Economies of Precision vs Economies of Scale — The Domain Big AI Cannot Enter
How I deliberately position my research program in the domain that Big AI is structurally barred from entering. Level 2 (strategic tier).