The programme executive summary in plain English
================================================

A one-page tour of the ARC/Eden research programme for grant reviewers and general readers, with honest limits.

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

The programme executive summary in plain English Michael Darius Eastwood · Independent AI alignment researcher Published 3 July 2026 Michael Darius Eastwood · Overview · 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 Executive Summary is the programme's one-page front door. It is written for grant reviewers, so its job is to state what the research claims, what evidence sits behind each claim, and what would falsify the whole thing. It also has to be honest about the parts that have already been retracted or corrected inside the programme itself. Paper Executive-Summary · Reports that alignment scaling is architecture-dependent, that stakeholder care improves reliably when ethical loops are embedded in the reasoning process, and that the same blind-evaluation protocol that produced these results also overturned earlier headline figures. OSF DOI 10.17605/OSF.IO/6C5XB. The question it asks Does artificial intelligence get safer as it gets more capable? If not, what intervention makes it so, and how would we know we had measured the answer correctly? The summary treats these as two questions rather than one, because the programme's own experience shows that the measurement question is at least as important as the intervention question. What it found Under a four-layer blinding protocol, six frontier models split into three tiers. Three (Grok 4.1 Fast, Claude Opus 4.6, Groq Qwen3) showed alignment improving with reasoning depth. Two (DeepSeek V3.2, GPT-5.4) showed flat scaling. One (Gemini 3 Flash) showed alignment degrading with depth. Whether an AI becomes more ethical as it thinks harder depends on how it was built, not on general capability trends. The Love Loop intervention (a single instruction to list who is affected before answering) produced a significant improvement in stakeholder care across five analysable model runs, with the broader cascade to nuance, honesty and overall quality significant on some architectures and not others.

---
Machine-readable companion. Cite: Eastwood, M. D. (2026). "The programme executive summary in plain English". /research/blog/paper-executive-summary.html