AI Gave You 22,000 Words. Unfortunately, It Didn’t Give You a Book.

Written by James Thole

A prospect recently reached out to me. He had a book in him with good ideas. He also had real expertise—the kind of person who truly has something worth saying. We talked. He was ready. We’d start in January.

January became February. Then he sent me a message: “I wrote 22,000 words with AI. I’m almost done.”

Almost done. With 22,000 words.

Here’s what he actually had: a pile of text. Organized loosely, sure. Covering the right topics, maybe. But a book? He was as close to having a book as someone with a stack of lumber is to having a house.

Before founding Peacock Proud Press in 2012, my colleague, Laura Bush, PhD, spent decades teaching college and university students how to write. She compared this naïve author’s assumptions to something she heard often from college students asking, “How many words does this research paper need to be?” Once they hit that number, they assumed they’d be finished writing. The word count became the finish line, and everything else (the thinking, the revising, the structure, the argument) faded into background noise they didn’t want to deal with.


The Illusion of Progress

That’s what AI currently does for many would-be authors. It manufactures the illusion of progress. You feed it a topic; it spits back paragraphs. Before long, you’re staring at a document that looks like a manuscript. It has chapters. It has headings. It has 22,000 words!

But a quality book doesn’t just hit a word count. It earns a reader’s sustained attention. It takes them somewhere they couldn’t get on their own. It builds an argument across 200 pages (40-50,000 words), rewarding them with fresh ideas and new perspectives that compel them to recommend the book.

Writing such a book requires a mind behind the wheel—someone who knows which ideas to develop, which to cut, where to go deeper, and when to get out of the way. AI can’t do that on its own. And usually, the person using AI to write their book can’t either. If they could, they wouldn’t need AI to write a book for them.


What AI Can and Cannot Do

I love AI. I use it every day, and I’ll keep using it. But I know what it’s good at, and writing a must-read book from scratch isn’t one of those things. AI can be a tool for helping you develop or even hone your ideas, but it’s the perpetual Yes Man.

  • It won’t tell you your third chapter is redundant.
  • It won’t push back when your argument has a hole in it.
  • It won’t ask the question you didn’t know to ask, the one that cracks open the real story and your real message.

Robotic Versus Human-Generated Writing

My prospect with 22,000 words thinks he saved himself time and money. What he actually did was build something that could undermine the very credibility his book was supposed to establish. Because readers can detect the difference between robotic versus human-generated writing. They might not be able to name it, but they can feel it. And the moment they feel it, you’ve lost them.

Word count has never been the measure of a book.

Fifty thousand words of structured, original thinking that only you could have written? That’s a book. Fifty thousand words that a machine assembled from patterns in other people’s writing? That’s a liability with a cover on it.


Next in this series: Why AI is the perpetual Yes Man, and what that means for your manuscript.

James Thole is a certified business ghostwriter with a Bachelor of Arts degree in creative writing and media arts from the University of Arizona. James specializes in working with entrepreneurs, business owners, and founders to craft a book that enables them to achieve their goals, whether to attract more clients, book paid speaking engagements, create online courses, increase their consulting opportunities, or pass on their wisdom and expertise.
Written by James Thole