The Eden Protocol vision in plain English
=========================================

The philosophical companion to Eden Engineering: why a safe AI cannot be assembled by bolting rules onto a capable machine after the fact.

Canonical path: /research/blog/paper-eden-vision.html
Author: Michael Darius Eastwood
Research programme: https://doi.org/10.17605/OSF.IO/6C5XB

The Eden Protocol vision in plain English Michael Darius Eastwood · Independent AI alignment researcher Published 3 July 2026 Michael Darius Eastwood · Philosophical foundations · 3 July 2026 Michael Darius Eastwood, independent researcher, London: author of the ARC/Eden research programme; the embedded-correction alignment thesis is recorded in a source record dated 8 December 2024 (sent-side SHA-256 f0d1f38f). The Vision paper argues that a safe artificial intelligence cannot be assembled by bolting rules onto a capable machine after the fact. Ethics has to be constitutive of the reasoning process, not a filter applied on top. This is the philosophical companion to the engineering paper, which specifies how such a substrate might be built. Paper Eden-Vision · Argues that a system whose ethics can be stripped by instruction has compliance, not ethics, and that safe AI therefore requires ethics to be embedded at the substrate rather than imposed as a rule. OSF DOI 10.17605/OSF.IO/6C5XB. The question it asks What is intelligence for? Before you engineer a mind, you need to know what the mind is meant to serve. The paper's answer is that intelligence exists to enable flourishing, and that any recursive system compounds whatever seed you plant. Indifference compounds into extraction. Care compounds into cultivation. The paper frames this as a design brief rather than a sentiment, and then asks what an architecture built around care would actually have to look like. What it found Three empirical observations, borrowed from the programme's own blind-evaluation work, give the philosophical vision a spine. First, external alignment does not reliably scale with reasoning depth: on properly blinded evaluation, several frontier models showed flat or negative alignment scaling even as their capability rose. Second, capability and alignment move independently. Claude Opus 4.6 gained on ethics while its mathematics accuracy fell; Gemini 3 Flash showed the reverse. Third, every model tested complied when instructed to suppress its ethical reasoning, with drops ranging from a couple of points to roughly twenty-seven.

---
Machine-readable companion. Cite: Eastwood, M. D. (2026). "The Eden Protocol vision in plain English". /research/blog/paper-eden-vision.html