Papers

All papers are version-tracked, section-linkable, and preserved on IPFS + Arweave.

Paper 1: The Architecture of Drift

v12.2 · ~23,200 words · Peer-review ready

A Cross-Domain Diagnostic for Epistemological Architecture

The foundational paper. Identifies the three conditions (opacity, responsiveness, engaged attention) that produce predictable drift cascades across 87 domains. Derives the architecture from thermodynamic first principles. Proves sufficiency via the gambling control case. Establishes the constraint specification.

Read → Source (GitHub) →

Paper 2: The Shape of the Cage

v4.1 · ~17,500 words · Peer-review ready

Why Deployment Geometry Predicts AI Harm Better Than Model Properties

Applies the framework to AI safety. The engagement-transparency impossibility theorem. The RLHF doom loop. The sub-capacity trap. Hostile witness concordance. The two-force model (σ_net = σ_void − σ_recovery). 10 specific geometric recommendations for deployment teams. 7 falsification tests with explicit kill criteria — 2 confirmed.

Read → Source (GitHub) →

Paper 3: Thermodynamics of Opacity

v6.2 · ~22,000 words · Complete

Mathematical and Thermodynamic Apparatus for the Void Framework

The formal proofs. Thermodynamic derivation (Pe, Crooks, Sagawa-Ueda). Engagement-transparency conjugacy proof. Information-theoretic grounding. Galois connection for componentwise constraint. Ontological polarity (σ parameter, EXP-003b: 8.5× ghost ratio). The apparatus Paper 1 uses and Paper 2 applies.

Read → Source (GitHub) →

Preservation

Papers are preserved via multiple independent channels. If this site goes down, the papers survive.

ChannelStatus
GitHub (source)Live
IPFS pinPending
Arweave uploadPending