The witch-hunt question is still “Did you use AI?”
It’s a yes/no that collapses a whole production into a verdict. Detector score. Deal canceled. Reputation on fire. We’ve watched that movie all year — manuscripts yanked after auctions, publishers admitting nobody actually knows how to police this, readers treating a percentage like a moral grade.
Today an indie author named Gregory Moulinet published something quieter and a lot more useful. He made a twelve-hour, full-cast audiobook of his novel with AI voice tools, spent close to a thousand dollars and four months directing the thing like a studio, and then he did the rare part: he wrote down where he delegated, where he refused, and where the machine put the creative weight back on him.
His replacement question is the one that matters:
How did you use AI?
Not undetectable. Accountable.
That’s the big world still stuck on the binary, and a little world starting to talk like grown-ups.
Moulinet’s piece is called “The Cost of Authorship,” and the cost is not just the ElevenLabs bill. It’s the interpretive labor that comes back when the “actor” has no lived emotion and infinite patience. He cast. He directed. He rewrote the manuscript as a performance script. He kept intention, selection, pacing, and moral ownership. He also says out loud that choosing a machine voice over a human performer is an ethical conundrum he isn’t pretending to solve.
Been there. And no shit — this is the conversation publishing keeps dodging while it builds detector pipelines and contract clauses that still can’t answer the only question readers actually need.
Circle is being built for that second question. Assignment. Review. Credit with a name on it. Disclosure by role — research assist, draft help, line polish, voice generation — not a purity contest and not a hide-the-tool game. A human reviews, edits, and stays accountable for the judgment.
If your process note can survive daylight, you don’t need a magic score. If it can’t, the detector isn’t your problem.
Write like someone will ask how. Then answer.
Also on the beat
- ChatGPT landed inside Microsoft Word as a sidebar for drafting, rewriting, and proofing without leaving the doc — the writing surface and the model now share one pane. https://runtimewire.com/article/openai-chatgpt-word-office-add-in-free-plans
- Publishers Weekly’s “Publishing’s AI Reckoning” maps the industry four years in: ops tools yes, creative trust still fragile, detection controversies, and licensing deals sitting next to lawsuits. https://www.publishersweekly.com/pw/by-topic/industry-news/publisher-news/article/101215-publishing-s-ai-reckoning.html
- UiPath CEO Daniel Dines co-wrote a 168-page book with Claude and ChatGPT as an adversarial Socratic dialogue — AI proposes, human decides, cutting is the real work. https://fortune.com/2026/09/10/daniel-dines-uipath-ceo-ai-book/
- Substack readers can run Pangram AI-likelihood scans on posts, with an optional author process note that quietly admits a score can’t capture “assisted.” https://techcrunch.com/2026/07/22/substacks-new-tool-tells-you-whos-been-writing-their-newsletters-with-ai/
- YouTube keeps rolling automatic labels for significant photorealistic AI video even when creators don’t self-disclose — platform labeling as the floor, not the whole story. https://blog.youtube/news-and-events/improving-ai-labels-viewers-creators/
How this was made
AI assisted research and a first draft of this Circle View. A human reviewed, edited, and is accountable for the judgment and the final take. Disclosure by role — not undetectable.
- A score is not a confession → Page 1 Still think a percentage is proof? Read Do AI detectors work on fiction?
- A Detector is NOT a Judge → Page 1 The short version of why a score can’t judge a book: Do AI detectors work on fiction?
- Prove you didn’t use a machine → Page 1 If someone’s asking you to prove a negative, start here: Do AI detectors work on fiction?
- Disclosure that never faces the reader → Page 2 Here’s how we keep it simple in Circle, no formulas: How Circle handles AI
- How did you use AI? → Page 2 We ask once, and your word does the rest: How Circle handles AI
- Put it on the Imprint Page → Page 2 Why I’d rather skip the label tiers altogether: How Circle handles AI
- Watermarks don’t sign your name → Page 1 Watermarks, detectors, same problem: Do AI detectors work on fiction?
- They knew the shelf would get crowded → Page 2 Slop is the problem, not the tool. Here’s where Circle stands: How Circle handles AI
- Label the voice → Page 2 How we handle AI in Circle, without the labels: How Circle handles AI
- Byline is no Costume → Page 2 Your name, your word, your work. That’s the deal in Circle: How Circle handles AI
