I wrote and published two books using my method to understand (and proof) the workflow. To make it short, you need to find a book structure that allows you to disconnect the chapters that you write in your book. What I mean by this is: The reader can jump into any chapter without the need to read the previous ones.

Think about a book about making money

You could start with the basics, e.g. budgeting, making more than you spend, re-investing the money that you make through investments for a compound effect. These ideas can be connected and stacked, obviously, but it is very hard for AI to "remember" what you wrote about personal finance in chapter two when you've reached chapter twenty-one that deals with Exchange Traded Funds (ETFs).

Therefore, organize your book in a way that chapter two explains personal finance in itself, while chapter twenty-one explains ETFs without referencing anything from chapter two.

For my tutorial case study I went even a step further: I wrote a book about how to be happy and I called it "99 Ideas To Think About If You Want To Be Happy"

To write it, I had ChatGPT generate 99 quotes from famous people about happiness and write an article of more than 500 words for each. The result was a 545-pages kindle book (480 pages as paperback). And as you don't need to know quote 45 to understand quote 76, the book never falls apart because AI "forgot" what quote 45 was about when it wrote the article for quote 76.

If you're interested in seeing the full process of generating the book, watch the video below as it details everything from start to finish: