What an AI Detector Score on Your Manuscript Is Worth
An AI detector looks like a breathalyzer and works nothing like one. It measures resemblance, not residue, and careful editing is what makes honest writing fail it.
Tag: AI Writing (Page 3)
AI WritingWhat It Costs to Fix an AI-Written Manuscript
The author thinks he is eighty percent finished. He is at page one, and somebody has to tell him. Here is what an AI-written manuscript really costs to turn into a book.
AI WritingWhen a Client Thinks the Ghostwriter Used AI
A publisher pulled a novel over work done by someone the author hired. If you have paid a writer and you are uneasy, ask for the trail instead of the score.
AI WritingThe Work You Would Never Have Started
Every productivity claim measures the same thing. The interesting category is the work nobody ever attempted, because the alternative was not a slower version. It was nothing.
AI WritingAI Detection Cannot Be Evidence, and Publishing Is Using It That Way
An author can now lose a publishing contract because a stranger uploaded their manuscript to a website and got a number back. Here is why those numbers cannot bear the weight being put on them, and what a writer should…
AI WritingAI Consciousness Left Philosophy and Entered the Laboratory
Researchers trained a model on a meaningless maze and found a positive and negative axis that already existed in the base model. Steering it changes behavior far outside the maze. Here is what that evidence shows, what it does not,…
AI WritingShe Bought a Manuscript an AI Wrote
A new client arrived with 45,000 words she had bought from an AI service. She still hired a ghostwriter, because the tool never supplies the material. The story was in her. The manuscript was around it.
AI WritingThe AI Op-Ed Argument Is Aimed at the Wrong Question
Stanley Druckenmiller used AI to write a Wall Street Journal op-ed and admitted it. The argument that followed was about disclosure, and disclosure was never the question. Ghostwriting settled this a century ago: authorship lives in having the thought, not…
AI WritingLinkedIn Built a Report Button That Pays the Person Clicking It
LinkedIn shipped a one-click AI slop flag with no definition, no training, no notification and no appeal. The structural flaw sits deeper than any of that. In a feed where attention is finite, suppressing a competitor makes room for you,…
AEOMarketing Does Not Have to Be Difficult. It Has to Be Scarce.
Alex M H Smith argues that marketing stopped working because everything got easy, and that the fix is to make yours harder, more tedious and more expensive. The diagnosis is right and the prescription confuses a proxy for the mechanism.…
AI TransparencyPublishing Is Cancelling Books Over AI. It Is Testing the Wrong Thing.
Publishing spent 2026 cancelling books over suspected AI use, on evidence that amounts to a detection score and a bad feeling about the prose. Detection tools measure polish, and polish is what a professional edit produces. Three questions would do…
AIThe Claude Watermark: What It Does and Why the Panic Is Wrong
Anthropic started watermarking Claude's text output on August 2, 2026, and almost every piece of coverage got the mechanism wrong. Nothing was added to your writing. What the mark does, what it proves, and why the SEO panic collapses on…