My AI usage
How I use generative AI
Published on 2026-03-19
I do use AI, but I feel I need to disclose how, to avoid giving the impression that I only use AI.
I have used AI on multiple fronts. Here they are, in no particular order.
Image generation
When I learned that a friend was launching a naturist webcomic portal, I wanted to contribute as a writer. I couldn't contribute otherwise, as I can't draw to save my life.
But I don't know how to write for a web comic!
So, I used an AI image generator to create the images for an attempt at a webcomic, Amanda in Love.
It became moderately popular, so I continued, but my goal wasn't to become an AI webcomic writer; it was to become a comic writer who would partner with an actual artist.
Only, in the process, I realized I didn't like writing webcomics. I don't have the knack for them. The portal was made, and the comics were so much better than what I could do.
I did use one of the rejected images for https://naturist-books.com/ and the logo for https://naturistcreators.com/ Was made by AI using very precise instructions.
Website Coding
Most of the code for the pending Naturist Resort Simulator game was made with the assistance of AI. I didn't just ask AI "Make me a game". I would type in full details the prompt for a portion and often edit the output.
But most of my websites were actually coded by hand.
Generative text
The entirety of the Books of Christopher was written by AI and was disclosed as such. And most of the chapters of that series were written in plain English and rewritten by AI into a gospel-like format. Many of the interesting quotes are actually mine.
The tracking of the language of the Weird visitors was attempted with AI but abandoned since ChatGPT was unable to keep track of the syntax rules.
That's it! No other episode, blog, or text was created by AI, but some words are from AI... as explained below
AI assisted grammatical revisions
My daughter was officially diagnosed with a learning disability that affects her ability to write properly. I am convinced I have the same one, but it was too late for me...
AI is used in two ways for editing and revising the texts:
- Via a browser plugin that highlights typing errors, notably using AI for it and for the recommandations
- Prior to publication, the texts made from late 2025 on are all sent to ChatGPT to review the quality of the writing.
The former tool will replace mistyped words or wrong-tense verbs with automatic or AI-generated corrections. This is not voluntary use of AI and the tool doesn't disclose when AI is used.
The latter, however, will often propose new wording for weird sentences, despite not being asked for them. Sometimes I use the new phrasing, therefore technically adding AI-generated content in my texts, but in that case, only for language improvements. The way the process is used is that ChatGPT is instructed to indicate the error, the correction, and why it is proposing this. It is not asked to restate the episode in better English and all corrections are made one by one.
AI assisted formal content revision
In some cases, the story occurs in front of a judge, or with social services or a doctor, or other officials. A first draft is always written, and ChatGPT is used to make the scene more rooted in reality.
In this case, AI is used as a sort of last-minute fact-checking. Occasionally, ChatGPT will volunteer a formal sentence that a judge or a doctor is likely to say, and will be used as if we had googled "How would a judge phrase this"
But the scene in general is written by us.
AI assisted legality check
My Life So Far and Overcoming Trauma has naturist children. The law limits what we can write about minors in a naturist setting.
For example, in one episode, Julie in the story mentally reacted to the size of the breasts of one of the naturist girls she saw, and in another case, to the pubic hair of another.
This is legal if both the narrator and the person described are adults, but not if they are minors.
ChatGPT was used to ensure not only that the texts respect the law but also that most stories fit in PG-13 rules, and when they break them, it's usually not because of sexual content. It could be due to language or violent content.
I am not a lawyer, and technically, ChatGPT isn't one either, but it was able to help keep a few stories cleaner, as in no case did we want to break the rules.
AI Assisted continuity check
When you post all the chapters of a story in a single ChatGPT tab, you can ask it for a continuity check. Like "Did I use the wrong name?" or "Do I have the right tone for this secondary character?"
When a story was put aside for a long time, it is useful because while I might not have worked on a story for 3 months, ChatGPT didn't forget the past episodes.
This usually doesn't cause more AI content to be added to the stories, but it might propose a suggestion that might be reworded to fit better. Often a single sentence. Like
"Cassian wouldn't say this; he would say that instead", and I will go, "Right".
In reality, this is rarely used, and when it pops up, it's usually when ChatGPT decides on its own to offer a suggestion.
And to be fair, it often flags out of character moments, which are voluntary. Like when a character is wounded emotionally and lashes out at a friend. "Marge would never say that" it might say, wrongly.
I just ignore it when that occurs, except in the 1% of cases where it's right and I rephrase.
AI assisted planning
This is kept for the end, for a good reason. When you post an episode, ChatGPT will often propose what the next step is, and 95% of the time, it's wrong!
But 5% of the time, it will suggest that I am going too fast or that two chapters in a row would benefit from a breather between them.
The next chapter will be written by me, but the impulse to write it, was from AI, so I am self-reporting it.
Every time I asked ChatGPT what the plan for the next 10 chapters or so could be, the result was always catastrophic.
If you ask ChatGPT to write a murder mystery, it will end in the 3rd chapter.
And yet, sometimes, it will suggest taking a breather or splitting a chapter, or keeping a reveal to a few chapters down the road. When I listen, AI affected the writing.
Conclusion on AI use on writing
Even if AI is used a lot, not a lot of the writing is actually affected by AI. A typical My Life So Far or Overcoming Trauma episode will have perhaps 3% of AI content, as follows:
Out of 2000 words, that's 60 words.
Of those 60 words, 40 were corrections by the AI part of the grammar tool.
Of the remaining 20, 15 were word substitutions or verb tense fixes from ChatGPT
The remaining 5 cover the few episodes where ChatGPT replaces a whole sentence, and I accept it.
What AI isn't used for
Writing (other than the Books of Christopher and rare rephrasing), idea generation, character creation (at all), plot generation (at all), and story planning (other than direct feedback on pacing).
The various games you will find in Mrs. Sanchez's classroom were either games my daughter or I had in class, or that we read about or invented on our own (the dice game).
The house design for Overcoming Trauma was designed by me, and the interview with the environmental engineer was written by my daughter, with no AI input at all.
All the plot twists in all the stories are either directly inspired from real events (like the delayed puberty of my daughter) or are created by me (or my daughter).
My sites are not AI slop content, and never will be, unless for a special project (like the books of Christopher, which are explicitly labeled as such).
Other Blogs
You can read the previous author's blog on naturist-fiction.com below. Blue Sky, Patreon and Reddit are the best way to get in touch https://bsky.app/profile/naturist-fiction.bsky.socialOld Blog
Get in touch
