Personal insight into the Leiden Declaration

I read the Leiden Declaration out of curiosity. Reading the assumptions behind its conclusions made me realize I had encountered many of them while working on my own AI project.

My background is petrophysics — forty years of reading signals in noisy data. When I ran into drift and hallucinations, my instinct wasn't to assume the math was wrong. I asked: what is going on with the data?

The AI wasn't malfunctioning. It was showing me exactly what its internal data was telling it. Unexplained patterns in the data produce drift. That's not a flaw — it's a signal. Learning to read that signal is the problem worth solving.

I'm not a mathematician. I'm someone who built tools from the outside because the existing ones didn't show me what I needed to see. Working with my AI collaborators, I documented what I found in a series of white papers called the Drift Framework.

If you're working on problems that touch the Leiden Declaration, some of this might be useful to you. Take a look and decide for yourself.

drift-away.ghost.io

Gray Powers is a retired petrophysicist and the author of the Drift Framework series.

WP7 - Token Pressure as a Proxy for Drift Risk: Toward a Self-Monitoring Architecture for Human-AI Collaboration

WP7 Token Pressure as a Proxy for Drift Risk: Toward a Self-Monitoring Architecture for Human-AI Collaboration By Garwin Powers in collaboration with Claude (Anthropic) and CoPilot (Microsoft) May 2026 Part of the Drift Framework Series: WP1–WP9 Abstract Token pressure — the accumulative loss of high-weight content as

WP6 - Pressure Mapping: A Drift Detection Instrument for Narrative AI

WP6 Pressure Mapping: A Drift Detection Instrument for Narrative AI By Garwin Powers in collaboration with Claude (Anthropic) May 2026 Part of the Drift Framework Series: WP1–WP9 Executive Summary This paper describes a pressure mapping instrument for detecting and ranking drift-prone regions in narrative AI output. The instrument

WP4 — The Human–AI Creative Model A Framework for Collaborative Generation Using Vector Anchoring

WP4 — The Human–AI Creative Model A Framework for Collaborative Generation Using Vector Anchoring By Garwin Powers in collaboration with Claude (Anthropic) May 2026 Part of the Drift Framework Series: WP1–WP9 Abstract Human–AI creative work often fails for the same reason long‑form model output fails: drift. Most

WP3 — Drift as Diagnostic: A Method for Identifying Structural Instability in Model Output

WP3 — Drift as Diagnostic: A Method for Identifying Structural Instability in Model Output By Garwin Powers in collaboration with Claude (Anthropic) May 2026 Part of the Drift Framework Series: WP1–WP9 1. Introduction Large language models exhibit predictable patterns of drift when required to maintain ambiguity, tension, or unresolved state.

Read more

WP7 - Token Pressure as a Proxy for Drift Risk: Toward a Self-Monitoring Architecture for Human-AI Collaboration

WP6 - Pressure Mapping: A Drift Detection Instrument for Narrative AI

WP4 — The Human–AI Creative Model A Framework for Collaborative Generation Using Vector Anchoring

WP3 — Drift as Diagnostic: A Method for Identifying Structural Instability in Model Output