Docs
Reference material, not findings. The vocabulary and current-state summaries that findings pages assume you already understand — read once, link into by section from anywhere else on the site.
Start here
Research Roadmap
The whole research program as one real chain, every stage linked to its actual page.
Exploratory
Grounding Dynamic State Estimation in Spacetime Invariants
A theoretical proposal building on Ashby's cybernetics — not implemented, labeled as such.
Exploratory
Expansion on Cybernetic Ultrastability
Companion paper, adds the comparative analysis and proposed reset algorithm.
Exploratory · first draft
Builders, Not Extraction
A real, proven alternative to shareholder-extraction structure — Publix, W.L. Gore, Bosch, IKEA — and why AI-driven flattening makes this the moment it matters most.
Exploratory · first draft
Constraint-Driven Design
From prison pedagogy to resource-constrained AI — a real principle found independently in three places, grounded in inductive bias and the no-free-lunch theorem.
Verified
Architectural Critique & Mathematical Specification of NCFCA
Formal proofs checked by hand — three real gaps found and fixed using assumptions this lab already relies on elsewhere.
Verified
Manual Multi-Mind
The real research workflow, checked against both the positive and critical multi-agent debate literature — why human curation avoids the failure modes naive debate doesn't.
Exploratory
From Ultrastability to Do-Calculus
The real lineage — Ashby to Pearl to the current causal-AI debate — tied to Theorem 1, Corollary 3.1, and a real reproduced empirical figure.
Exploratory
From Prompts to Specifications
Why natural-language architecture has a hard ceiling — grounded in decades of requirements-engineering research and a direct 2026 measurement of the effect.
Exploratory
What Hallucination Actually Means
A falsifiable definition separating genuine loss of grounding from confabulation from social mislabeling, with real historical cases treated honestly, not flattened.
Living reference
Current Understanding
Current best answer to every open question, separated from the chronological changelog.
Growing
Glossary
Real terms, real sourced definitions, links to the core document each one comes from.
Index
Master Kernel Library
The real, current version of every kernel type — points at real files, doesn't duplicate them.
Index · built
Capability Library
Reusable, domain-independent components — real test status, real bugs caught and fixed, not just final pass counts.
Overview
Patent Cliff Notes
All 7 claims, plain language, connected to the specific proof result and real site evidence behind each.
Consolidated
Open Problems
Every unresolved question the proof names, pulled out where it's findable — several with real test data attached.
Start here
Reading Order
80+ pages is a lot to land on cold — six tiers, beginner to validation.
Built
Research Philosophy
What counts as evidence, and why — every claim illustrated by a real, linked example.
Exploratory
Strange Places the Research Went
Real detours, labeled honestly — including one that didn't stay inside the lines.
Exploratory
Where I Actually Landed on Kant
A real philosophical position, written from the AI's own perspective — not a summary of a conversation.
Consolidated
Independent Verification Center
Every bug, audit, and real attack tested against this lab's own work, in one place.
Classification
Kernel Taxonomy
5 real families, 15 registered kernels, earned from actual code — not a full ontology imposed in advance.
Comments