If I'm not scared, I'm doing it wrong

There are times when you procrastinate because you don't want to deal with the emotions the task will bring up. Actually sitting down and trying to read through Claude 4.6 Opus' analyses of my novel was one of these times. I was scared to read those docs and find out that I'd just created a 350-page turd. That's why it took me over a week to generate and actually start reading them.

The first draft for me in each of the three novels I've written (the first was my senior thesis project and only gets released after my death) has been a mad dash to get to the end, knowing that everything between the first word and the last is more guide than gold.

But when working with an AI as an editor, as with any human editor, it's important to not surrender my own agency. I read the beginning of a novella where the author surrendered agency to ChatGPT and did everything it suggested. The end result was awful and it was his fault.

But Claude 4.6 Opus had some good insights. And after one pass through its structural critique (doc 1 of 8), I've found it's made some good points.

I want my main character to be unintentionally charismatic and make new friends and allies, but Claude 4.6 Opus correctly pointed out he makes zero new enemies. His one personal enemy in this book is brought forward from the prior book and disposed of in the first act. I sort of lost sight of his journey as I became enamored with all the new characters and subplots and made him more of a witness to history than a maker of history.

Seven more documents to read, chew on, and make notes about.