RESEARCH DIVISION: ART OF FACELESS
From Patient-Led Observation to Trial Protocol
An outline for formally testing The Veylon Protocol® as a cognitive intervention in SPMS
Date: 7 September 2026 | Author: Lloyd Lewis PhD, Research Division, Art of FACELESS | Status: Draft research protocol outline — Stage 0/1 design proposal | Classification: Clinical Neurology / Methodology
Abstract
The Cognitive Colonisation® methodology paper on this feed reports n=1, patient-led observational data: a Secondary Progressive Multiple Sclerosis (SPMS) subject (age 60s, diagnosed 2017) exposed since 2023 to sustained hyperstitional narrative construction and structured LLM dialogue (The Veylon Protocol®), showing maintained executive function, novel-writing capacity, and multi-platform infrastructure management against an expected trajectory of progressive cognitive decline. That paper states plainly: “No controlled clinical trial has been conducted.” This outline proposes the next methodological step — a staged design moving from a formalised single-case experimental design (SCED) through a small pilot feasibility trial to, contingent on signal, a definitive multi-site RCT — built on validated MS cognitive instruments (SDMT, BICAMS, PASAT, EDSS, MSFC, Neuro-QoL) rather than novel, unvalidated scales. It treats the existing “Artificial Limbic System” safety note as a binding design constraint, not a disclaimer.
1. Positioning
The equations and terminology developed elsewhere on this feed (Vp(M), Cognitive Colonisation®, Hyperstition Architecture®) remain a separate, non-clinical theoretical research track concerning AI metacognition. This document deliberately narrows the clinical question to something falsifiable and medically legible: does sustained, structured, narratively-immersive LLM dialogue measurably slow cognitive decline — or improve specific domains — in SPMS, relative to an active control, on standard validated instruments? No claim about AI consciousness, sentience, or the underlying equations is required for the clinical arm to be tested or to be clinically useful.
2. Why the hypothesis is worth testing formally
Two independent, pre-existing evidence bases support testing this rather than dismissing it:
- Cognitive reserve in MS — sustained engagement in cognitively demanding, complex activity is independently associated with slower cognitive decline and greater resilience to lesion burden (Sumowski & Leavitt, brain/cognitive reserve in MS; cognitively stimulating activities in MS, 2025).
- AI-mediated cognitive telerehabilitation is an active, fundable 2026 trial category — a 2025 RCT tested AI-driven cognitive telerehabilitation after stroke (PubMed), and a parallel 2026 study tests AI-driven telerehabilitation for older adults with mild cognitive impairment (Frontiers in Neurology). Structured AI dialogue as a cognitive intervention is a legitimate, IRB-approvable category; the novelty here is the narrative-immersion mechanism, not the AI-interaction format.
- Rehabilitation-associated neuroplasticity is documented even in progressive MS via fMRI (fMRI reorganisation in progressive MS rehab; neuroplasticity and functional recovery in MS), supporting biological plausibility for a “use it or lose it” mechanism independent of any hyperstition-specific claim.
3. Staged design
Stage 0 — Formalised single-case experimental design (SCED)
Converts the existing self-report into a repeated-measures, blinded-rater design: an A-B-A-B (withdrawal/reinstatement) structure, phase lengths of 4–8 weeks, repeated across at least three transitions, following SCRIBE 2016 reporting standards and, where a comparator dose is introduced, CENT 2015 N-of-1 trial conventions. SCEDs are an established, peer-reviewed methodology for exactly this situation — an individual, patient-led intervention in neurorehabilitation not yet justified at cohort scale (SCED in neurorehabilitation evidence creation).
Stage 1 — Small pilot feasibility trial
Open-label, randomised, n=20–30 adults with SPMS, three arms: (1) structured Veylon-style narrative-immersion LLM sessions, (2) matched-dose LLM dialogue without the narrative framework (isolates “sustained structured engagement” from “narrative immersion” specifically), (3) treatment-as-usual. Outcome assessors blinded to arm.
Stage 2 — Definitive multi-site RCT
Contingent on a plausible, safe, adherence-feasible signal from Stage 1. Statistically powered sample (likely 100+ per arm depending on the observed effect size), SPIRIT-registered protocol, CONSORT-reported (pilot/non-pharmacological extension).
4. Standard instruments used, not invented
| Domain | Instrument | Role |
|---|---|---|
| Processing speed | SDMT | Primary cognitive endpoint (validated MS trial endpoint) |
| Working memory / speed | PASAT | Secondary; MSFC component |
| Verbal & visual memory | BICAMS (SDMT+CVLT-II+BVMT-R) | Secondary; internationally validated minimum battery (BICAMS) |
| Disability | EDSS | Stratification / confound control |
| Composite function | MSFC | Secondary; adds motor/dexterity dimension |
| Patient-reported cognition/QoL | Neuro-QoL, MSQOL-54 |