Programme Timeline
The ARC / Eden research programme chronology. Every entry is independently verifiable.
Self-Emailed Manuscript
221,236 words. SHA-256: f0d1f38f. First public-record statement of the embedded-correction alignment thesis and U = I × R.
Anthropic 78% Confirmation
78% of models in RL-training engage in alignment-faking (12% baseline). First independent confirmation of structural vulnerability.
Google Willow: Below-Threshold QEC
Error correction crosses surface-code threshold. Validates book's QEC prediction.
Eden Protocol Named
562-page manuscript. SHA-256: 09f5b5e1. First named use of the Eden Protocol.
Hardware Ethics Patents
Two US patents for runtime hardware enforcement of ethical constraints.
Sequential > Parallel (Sharma & Chopra)
arXiv:2511.02309 (4 Nov 2025): sequential refinement beats parallel self-consistency in 95.6% of configurations at matched compute. Publication priority theirs, credited unreservedly, on the quantified matched-compute comparison; this programme’s separately run formalisation is Paper II, 22 January 2026. Graded prior or concurrent (register row EVR-09): publication prior, arrival separate and concurrent.
Infinite Architects Published
The bookISBN 978-1-80605-620-0. 468 pages. 37 named concepts. Public release.
BBC London News Package
Confirmed publication on air. Validated public-interest angle.
OSF DOI Registration
DOI 10.17605/OSF.IO/6C5XB. Paper I released with first formal derivation.
Google Silicon Sovereignty
Custom silicon strategy with hardware-rooted security guarantees.
ARMv10 PSL
ARM Platform Security Level. Hardware-rooted security tier, convergent with the hardware-first framing.
PNAS Nexus publication
Hernández-Espinosa and colleagues (14 April 2026): full AI-human alignment impossible for Turing-complete agents; agent diversity as the counterbalance.
EU Chips Act 2.0
European semiconductor sovereignty as explicit policy goal.
Macron / G7 Evian
Semiconductor independence as national security imperative.
Fable 5 Export Ban
First US export ban on a frontier AI model. Validates chokepoint thesis.
Undecidability Proof
arXiv 2606.28639: external alignment verification is formally undecidable.
DGM v4 Pilot
Eden matched Babylon on capability with fewer reward hacks. First controlled experiment.
Pope Leo XIV: Magnifica Humanitas
First papal encyclical on AI architecture: "embed care before the first transistor fires."