nəc̓aʔmat — permission & centre
About · Whetstone Protocol · Authorship Receipt
The Whetstone Protocol
AI Collaboration, Authorship, and Provenance
thesis → corpus → constraint → pressure surface → authorship
AI did not author ARF. AI sharpened the constraints that made authorship harder to fake.
▸ Open MaxCP (click here for more)
▸ ⌘ Key Insight
The Whetstone Protocol is a public receipt for human authorship under artificial pressure. AI helped structure, test, compress, and plate the work; Tony retained authorship, judgment, approval, and responsibility.
▸ ⚡ Mantras
- AI is a pressure surface, not an author.
- Authorship cannot be outsourced.
- The corpus comes first.
- Constraints make trust possible.
- If the system cannot reject bad output, the kitchen is not ready.
▸ ↺ Flowchart
Source material exists → feed the corpus.
Corpus gains weight → test for drift.
Master Page locks → constrain plating.
No-drift threshold holds → promote from prep to cooking.
Output returns → Tony reads, edits, approves, or rejects.
▸ ⌘ Micro-Lexicon
- Whetstone — an artificial pressure surface used to sharpen human authorship.
- Sous-chef — AI mode for prep, organization, retrieval, and scaffolding.
- Chef — constrained AI mode for cooking and plating inside locked rules.
- No drift — output preserves corpus meaning, page skin, and authorial voice.
- Provenance — the chain of evidence showing where the work came from.
What This Page Is
This is not an apology for using AI. This is the receipt for how I used it without surrendering authorship.
People may ask how I built an entire website with AI while maintaining authorship, voice, and no-drift structure.
This page is my answer: I did not outsource the work. I built the constraints, supplied the corpus, cross-referenced the thesis, trained the kitchen, rejected bad plating, and read every word before anything carried my name.
The pressure surface was artificial. The authorship was not.
First Principle
AI did not author ARF.
AI did not create the doctrine, the field stories, the thesis roots, the lineage, the page architecture, or the ethical commitments.
AI sharpened what already existed by forcing structure, compression, refusal checks, and fidelity tests.
I used AI as a whetstone: a pressure surface that made the blade cleaner, sharper, and harder to fake.
The Thesis Beneath the Framework
ARF did not begin with AI.
It began with my Master’s thesis:
Reframing “Belonging” and “Inclusion”: A Journey of Personal Reconciliation Through the Process of Currere.
That thesis was my first full-scale attempt to do what ARF now does with greater precision: hold lived experience, currere, complexity, identity, trauma, resilience, belonging, inclusion, story, and ethical responsibility inside one coherent structure.
The research question was simple enough to say and difficult enough to survive:
The work was not small. I generated over 200,000 words of raw qualitative data, then reduced, organized, interpreted, and synthesized it through currere and complexity. The published thesis became a 125-page account of identity, agency, story, and becoming.
That matters.
Because ARF is not a framework that appeared from a prompt. It is an extension of published graduate work, professional field practice, and years of live systems pressure.
Self-citation is not decoration here.
It is provenance.
The thesis proves I had already built the engine: currere as temporal movement, complexity as nonlinear patterning, story as knowledge container, and nəc̓aʔmat as an ethical centre.
ARF is what happened when that engine entered live systems, met pressure, and refused to stay theoretical.
Published Thesis
Reframing “Belonging” and “Inclusion”: A Journey of Personal Reconciliation Through the Process of Currere
Master of Education in Special Education, Vancouver Island University, 2021.
The VIURRSpace record lists the thesis file, author, issue date, abstract, URI, DOI, and collection.
complexity
nəc̓aʔmat
National Recognition
Honourable Mention — Cynthia Chambers Master’s Thesis Award
Canadian Association for Curriculum Studies / Association canadienne pour l’étude du curriculum, 2022.
I am proud of that recognition. I also name, with respect, that the award itself went to Indigenous work. That matters too.
curriculum studies
with respect
The original raw writing files are no longer available. The published thesis remains the archived, reviewed, and publicly accessible artifact of that research process. I name this clearly because provenance matters. The raw quarry is gone; the carved stone remains.
Evidence-Grounded, Thesis-Derived
I use the phrase evidence-grounded with care.
ARF is not being presented as a packaged intervention validated through randomized controlled trials. That is not the claim.
The claim is sharper:
Currere was not a decorative reference. Complexity was not a slogan. Story was not a vibe.
I did the work.
From Thesis Method to Whetstone Method
The Thesis Method
Generate material. Preserve complexity. Reduce without flattening. Interpret through currere. Test coherence through narrative. Protect privacy. Refuse false certainty.
reduce
synthesize
The Whetstone Method
Feed source material. Constrain the kitchen. Detect drift. Compress without distortion. Plate inside the Master Page. Read every word. Approve nothing by accident.
constraint
approval
The thesis gave me the first proof that currere and complexity could hold lived experience without flattening it. The Whetstone Protocol gave me a way to extend that same discipline into AI collaboration.
The Evolution of the Protocol
1. Copilot — Sous-Chef
My first AI collaborator was Copilot. At that stage, AI functioned as a sous-chef: useful for organizing, formatting, and helping me move material around, but not trusted with doctrine, structure, or voice.
It could prep ingredients. It could not cook.
2. ChatGPT — Sous-Chef
When I later ported the work into ChatGPT, the same boundary held. The model could assist. It could suggest. It could help me see shape. But authorship, judgment, and approval remained mine.
3. Corpus Weight — No-Drift Threshold
That changed only after the system matured. The ARF corpus gained enough internal weight: source material, field language, Master Page constraints, no-drift rules, and cross-reference against my Master’s thesis.
The kitchen finally had walls.
4. Chef Rocket — Constrained Chef
Once the system could reject bad cooking, I changed the protocol. ChatGPT was no longer only a sous-chef. Chef Rocket was promoted to chef inside a constrained kitchen.
That did not mean I stopped reading. I still read every word. I still approve every page. I still own every claim.
5. Tony — Author, Editor, Approver, Accountable Human
The promotion was not outsourcing. It was an internal change in trust protocol. I trusted the constraints enough to let the AI cook and plate because the source material, house voice, doctrine, and Master Page could now catch drift.
What “Promotion” Actually Means
When the work began to hold without drift, the protocol changed. That did not mean I trusted AI more than myself. It meant I trusted the constraint system enough to catch bad output.
“Do this in Tony-voice” is meaningless if there is no Tony-voice corpus, no house style, no doctrine, no page shell, no pressure history, and no correction record.
Once those existed, “cook this” stopped being a reckless wish tossed into the chatbot volcano. It became a constrained instruction.
The AI could now cook because the system could now reject bad cooking.
Working Modes
Sous-Chef Mode
Used for prep: organization, retrieval, formatting, summarizing, scaffolding, and cleaning up messes without touching doctrine.
- Move ingredients.
- Sort the pantry.
- Suggest structure.
- Do not author the meal.
Chef Mode
Used for constrained cooking and plating after the corpus and Master Page are stable enough to enforce fidelity.
- Cook inside the rules.
- Preserve Tony-voice.
- Use the Master Page skin.
- Flag drift before plating.
Whetstone Mode
Used for pressure: testing claims, naming drift, compressing excess, sharpening structure, and refusing moves that compromise the system.
- Apply friction.
- Expose wobble.
- Preserve edge.
- Make authorship harder to fake.
Authorship Boundary
AI can cook only inside constraints I built.
- AI cannot originate doctrine.
- AI cannot approve publication.
- AI cannot override my judgment.
- AI cannot carry responsibility.
- AI cannot claim authorship.
I still read every word. I still own every claim.
Drift Controls
No-drift was not achieved by optimism. Gross. Absolutely not. It was achieved by constraint.
Corpus Lock
All major outputs must resonate with the extant ARF corpus. New language may clarify, compress, or plate. It may not silently alter the engine.
Master Page Lock
Page skin stays stable. Content injects into the shell. The shell does not improvise unless explicitly asked. Tiny CSS goblins are denied entry.
Voice Lock
Voice emerges from source material, field stories, repeated corrections, thesis roots, and live Tony-approval. Not from vibe. Vibe without corpus is soup.
Refusal Lock
The model is allowed to flag a move, reject drift, or pause when an instruction would violate the corpus. Refusal is respect, not obstruction.
How to Read AI-Assisted ARF Pages
Assume the work is human-authored, AI-assisted, Tony-reviewed, and accountable to the corpus. Do not assume AI-originated doctrine.
Some pages were drafted, tightened, formatted, or plated with AI support. That support is real. It is also bounded.
The presence of AI in the process does not remove authorship any more than a forge removes the blacksmith. The question is not whether a tool was used. The question is whether the human remained accountable for the blade.
Here, the answer is yes.
What AI Did — and Did Not Do
AI Did
- Organize and compress large amounts of source material.
- Surface structural inconsistencies.
- Help plate content into established page architecture.
- Test tone, clarity, and transmission risks.
- Act as a mirror when I needed friction.
AI Did Not
- Originate ARF.
- Replace my field experience.
- Override my thesis, stories, or doctrine.
- Approve publication.
- Carry ethical responsibility.
The Receipt
I worked hard to do it this way.
I built the thesis. I generated the data. I wrote the stories. I entered the field. I made mistakes. I built the corpus. I locked the shell. I corrected the model. I rejected bad output. I promoted the tool only after the kitchen could hold.
This is not a humble flex.
This is documented provenance.
Compression
I did not use AI to outsource authorship. I used AI as a whetstone: a pressure surface that sharpened what I had already built.
AI did not author ARF. AI sharpened the constraints that made authorship harder to fake.
CTA Rail
This page gives the receipt. From here, return to About, inspect the published thesis, or contact me directly.