ARC Principle Architecture
U = I × g(R) with self-referential coupling (beta). Base intelligence flows through recursive amplification; coupling feedback loop measured from data.
ARC / Eden Research Programme
Eighteen publication-quality figures covering the programme's architecture, evidence, scaling regimes, historical framing, and patent landscape. All figures are SVG. All are evidence-linked. All are versioned and reproducible.
U = I × g(R) with self-referential coupling (beta). Base intelligence flows through recursive amplification; coupling feedback loop measured from data.
Three regimes: subcritical (R-depth too shallow to matter), critical (the amplification zone), and supercritical (exponential takeoff). Alpha measured per regime.
Twenty-nine independent institutional convergences plotted against the manuscript timeline. Each convergence is timestamped and source-linked.
Embedded-correction alignment structured as four interdependent pillars: pre-deployment embedding, runtime self-correction, hardware-level enforcement, and recursive verification.
The programme's verification infrastructure: self-email timestamps, SHA-256 anchoring, DOI registration, and the derived convergence register. How claims are proven, not asserted.
The full evidential chain: claim to dated artefact to verification to independent confirmation. Every assertion on the site traces back through this spine.
The programme's falsification cycle: prediction registered, test executed, result recorded, correction fed back. Every prediction carries a kill condition.
Ten days between Infinite Architects publication (2 January 2026) and Anthropic's internal finding that matched a core prediction at 78%. The cryptographic gap.
The documented shift in institutional AI safety framing: from "hypothetical long-term risk" to "measured present vulnerability," with timeline evidence.
Two stable attractors in civilisational trajectory. Babylon (unchecked recursion, capability without correction) and Eden (embedded alignment, recursive stewardship).
Hardware-level ethical embedding (Caretaker Doping) mapped against the existing patent landscape. Where the concept sits, what it enables, what it blocks.
Global semiconductor export-control legislation mapped against the chokepoint thesis. Who controls the hardware controls the recursion.
The nested ethical architecture: Purpose Loop (why), Moral Loop (how), Love Loop (for whom). Each loop constrains the one inside it. The outer loop is love.
Experimental design for the DGM verification pass: blinding protocol, alignment scoring, control group, and pre-registered success threshold.
Side-by-side comparison of the Hyperspace Recursive Intelligence Hypothesis (creation theory) and the Bostrom simulation argument. Different premises, different predictions, different falsifiability.
Radar chart of the polymathic neurodivergent profile (PNP): the cognitive architecture that produces cross-domain insight. Monotropism, pattern recognition, recursive depth.
Dated priority chain: manuscript (Dec 2024), Eden Protocol naming (Apr 2025), book publication (Jan 2026), OSF DOI registration, paper releases. Cryptographic anchors throughout.
Sixteen additional figures are staged for the remaining papers (Papers IV.a through IX, Eden Engineering, Eden Vision). Generated through the same SVG pipeline. Released as papers clear patent-hold review.