TRANSPARENCY & METHODOLOGY
AI Attribution
As I lead engineering teams through the transition to an AI-native software development lifecycle (SDLC), I publish original findings, experiments, and lessons from production codebases. None of it is AI-generated slop; everything I share is grounded in direct experience.
For efficiency, I use AI across my workflow to organize notes, tighten prose, verify facts, and generate cover images. Because AI's role varies from piece to piece, new articles starting late 2026 carry a clear score so you know where the ideas and the words came from.
Prior articles remain unrated. The score runs on a 1-to-4 scale — higher scores mean more of the writing is human.
The Scale
I wrote it. AI did no more than proofread it or check facts.
The ideas, observations, and experience are mine. I talked them through with AI, usually as rough notes or dictation, and AI organized and polished the prose. Most of what I publish is here.
AI supplied the ideas and the first draft. I reshaped it, corrected it, and decided what to keep.
AI produced the ideas and the writing. I chose the topic, reviewed the result, and published it.
Reading the Line
The score describes where the ideas and the words came from, not how good the piece is or how carefully it was checked. Every article, at every level, is reviewed by me before it goes out, and I stand behind what it says.
When a piece mixes levels (for example, my own essay with an AI-written appendix), I score it by the main body and note the exception in the line.
What Is Always AI-Generated
Cover images on this blog are AI-generated art with my branding added on top. They don't affect the score.
Why I Disclose
I help engineering teams adopt AI, and I write about what works and what doesn't. Being clear about how I use it in my own work is part of earning that trust. It also lets you calibrate: a Level 4 essay is my voice end to end, and a Level 2 piece is closer to a well-edited briefing.
Where the Scale Comes From
There isn't a widely adopted standard for this yet. My scale is deliberately simple and borrows from two serious efforts:
a research prototype that walks a creator through how AI contributed to a piece and generates an attribution statement.
at the University of Waterloo, which breaks AI use down by role (ideas, research, analysis, writing, editing, visuals). It is licensed CC BY-SA 4.0.
If a credible standard emerges that improves on this, I'll evaluate it and update this page.
AI attribution: 3/4 — Human-originated, AI-shaped. What this means →
This policy page was conceived and directed by Jason Vertrees, with AI assistance for research compilation, code synthesis, and translation.