Member-only story
Using a Large Language Model to Workshop a Novel
Strictly personal and subjective observations on the process.
No text included here was generated by AI. The “quotes” below from ChatGPT are approximations.
I’ve been working on a novel since July, and as of last week, the draft is finished. Now comes what is for me the much less grueling task of editing — less so because I do a lot of the heavier editing as I draft (which I wouldn’t recommend for any writer who wishes to retain their sanity).
This is my second book, the first having been written in 2021 and ’22 and published in 2023. What makes this one different is that throughout the process, I used large language models (LLMs — Gemini at first, then ChatGPT), not to write the thing but workshop it.
I’ll reiterate that: an LLM didn’t write my book. If it had, it would’ve been terrible, for reasons I’ll try to explain below. Rather, I used it not as one tool, but several, including as a “reader” — another “pair of eyes” to lay on the text and provide feedback.
What follows is an initial sketch of my impressions of how it worked out — the drawbacks, benefits, and a few suggestions for improvement.
My favorite part about the MFA program I did was the opportunity to have my writing workshopped. I…
