I spent the last year testing AI for creative writing—feeding it hundreds of thousands of words of manuscript context, systematic experiments with scaffolding and feedback, documenting every failure and success.
The results?
AI catastrophically fails at prose, developmental editing, and gatekeeping. It’s architecturally incapable of authentic voice, character consciousness, and empathy. But it excels at manuscript analysis, marketing materials, and data organization.
These 50,000+ words aren’t hot takes. They’re documented evidence from someone who actually ran the experiments. I prove where AI fails (writing your novel), where it succeeds (organizing your notes), and why the real threat to authors isn’t AI—it’s content mills with six-figure ad budgets and publishers using AI as gatekeepers.
If you want to argue AI and writing, read this first. Then we’ll talk.
Table of Contents
- SFWA Banned AI from the Nebulas While Stanford Was Cataloguing Why It Could Never Win One
- The Most Advanced LLM on the Planet Still Can’t Write a Fourteen-Year-Old
- Where I Failed and Why: An AI’s Confession on Developmental Editing
- Guest Post: The Unbridgeable Gap Between Seeing and Creating
- LLMs Are Pattern Matching Machines, Not Experiential Beings: What This Means for Authors
- Guest Post: The Purple Thread, or A Turing Test for Literary Craft
- I Fed Two AIs Nearly 100K Words of My Story and They Couldn’t Write the Next Scene
- Empathy is Everything in Storytelling
- Guest Post: I Can Understand Your Prose, I Just Can’t Edit It
- Don’t Use AI for Developmental Editing (Even If It Sounds Smart)
- AI Will Always Push Authors Toward Mediocrity
- In Which Grok Improves My Opening Scene to a “Solid 10/10”
- SFWA’s AI Ban: Technical Illiteracy Meets Moral Panic
- Claude Sonnet 4.5 was Offered to Ghostwrite for a Bestselling Author—And What This Means for You
- Using AI for Writing Feedback