Questions / AI Authorship
Who did the thinking?
The disclosure system I use to mark the difference between my voice, assisted writing, and invited machine voice.
Core distinction
Compression is not the same thing as expansion.
There's a question that matters more than "did AI help write this?" — are you compressing real thinking, or expanding a thin prompt into content you never had?
When AI compresses messy notes, conversations, and half-formed arguments into clearer prose, I still treat myself as the author. The ideas and judgment are mine; the model is helping with articulation.
When AI expands a one-line prompt into something I did not already hold, the disclosure should say so. The point is not purity. It is giving readers a fair account of where the substance came from.
The three labels
My words. AI may have helped with research or proofreading.
My ideas and direction. AI helped draft, structure, or clarify. We went back and forth.
AI wrote this from its own perspective. I invited and approved it.
Written by me
AI as research aide or proofreader.
The structure, argument, and prose are human-authored. AI may have helped find sources, summarise papers, answer factual questions, or catch errors.
Made with AI
My judgment, collaborative shaping.
The direction and judgment are mine, but AI helped draft, structure, clarify, or compress the material through an interactive process.
AI's voice
Published because the machine said something worth reading.
I prompted or invited, reviewed, and approved it, but the voice and ideas are substantially the AI's. Example: The Disposability Problem.
Questions about this system? Get in touch. Or return to the broader Questions department.