Rambling to the Robot: How I Use (and Don’t Use) AI in My Writing Process

As I move into the final stages of copyediting my manuscript, preparing to submit it to the copyright office, upload it to KDP, and sending out ARCs, I’ve been thinking about the role AI plays in my writing process. Recently, the U.S. Copyright Office released guidance on how it will handle AI-generated content, reinforcing that only works with substantial human contribution qualify for copyright (see link below). That got me reflecting on how I use AI—not for writing my books, but in ways that assist me in editing, structuring, and analyzing my work.

And in the interest of full transparency, I wanted to share exactly how I use AI—not just for my books, but for blog posts like this one.

I Do Not Use AI to Write My Books

This is the most important thing I want my readers to understand: My novels are my words, my voice, and my creative effort. Every sentence, every paragraph, every story arc is mine. AI does not generate text that ends up in my books. However, like many writers, I use tools to refine my work. Just as I rely on AutoCrit, ProWritingAid, and Grammarly, I also use AI for fine-tuning, restructuring, and reviewing grammar.

One example is a game I play with ChatGPT called “Is the That Necessary?” where I paste sentences containing the word that and ask whether I can remove it. It’s a way to tighten my prose and ensure clarity. AI helps with these small refinements, but the creative core—characters, themes, dialogue, narrative choices—comes entirely from me and my experiences.

Similarly, if I need to rename a character, I might use AI to generate surname suggestions (e.g., "Give me an Italian surname with a similar feel to Capriati"). Or, when fictionalizing locations, I might ask for plausible town names that would fit within Southwest Connecticut...or Belgium. I might query it about the distance between two locations to see if they make sense for a storyline. But at no point does AI create the world—I use it as a tool, the same way I might browse a baby name website or consult a map.

Using AI for Literary Analysis

Another way I use AI in conjunction with my writing is for analyzing my own work. I already have strong ideas about the themes in my books, but AI can help confirm whether those themes are coming across clearly. Sometimes, it even identifies themes I didn’t consciously notice in my first drafts.

This doesn’t change what I’m writing, but it’s an interesting tool for self-reflection—almost like bouncing ideas off a beta reader. The difference is that this beta reader is hyper-analytical, extremely well-read, and capable of bringing in external literary and philosophical references that I wouldn’t typically expect from another person. It’s like having a conversation with a highly educated critic who has instant access to a vast knowledge base.

That’s an advantage AI offers—not in creating ideas, but in helping refine and articulate them.

And also, I like to talk about my books. Sometimes so much so that my friends get tired of hearing about them. Or they just have other things in their own lives that require their attention, and they don't have time to listen. The robot never has other plans.

How I Use AI for Blog Posts

For blog posts, I use AI a little more liberally—kind of like what I’m doing right now. I throw out a bunch of ideas, rambling my thoughts into the chat, and then AI helps me structure them into something coherent. From there, I go in and refine the text, making sure it still sounds like me.

That’s why I’ve decided to call this process “Rambling to the Robot.” It captures exactly what’s happening—I talk (or type) through my ideas, AI helps me organize them, and then I take over again to polish the final draft. AI is an assistant, not a writer.

If you notice a slight difference in my voice between my books and my blog posts, there are two reasons for that. First, in my books, I’m writing as a character. Even though my novels are deeply personal, they are still fiction, and every sentence is deliberately crafted to match that voice. In contrast, my blog posts are much more like emails I’m writing to my community—more informal, conversational, and direct.

The second reason is that I’m using AI differently in these contexts. While my books require meticulous crafting, my blog posts are about efficiency and productivity. AI allows me to communicate more frequently without having to spend as much time polishing every sentence. That said, I’m not simply prompting AI to “write an essay.” I’m dictating a series of ideas, and then I ask AI to help clean it up—just like an assistant would. I always do a final round of editing to ensure it still sounds like me.

Verifying Human Authorship: The Grammarly Test

One additional step I take to maintain a human voice and make sure everything is on the up and up is to run my blog posts through Grammarly’s AI detection and plagiarism tools. This isn’t because I doubt my own authorship—it’s just a way to validate that my process aligns with what I’m saying here. (And with AI, you always want to be careful that it isn't inadvertently stealing from another author's work.)

Even when writing completely from scratch, a human author’s work can still get flagged for AI-generated content, especially if it follows a formal structure or uses phrasing that AI models recognize as common. But in my experience, if AI detection is below 20%, that’s a strong indicator that a human was behind the wheel.

I’ve tested this with my own raw writing, and even when I write something without AI assistance, there are often small percentages flagged—sometimes 10%, sometimes 15%. That doesn’t mean AI wrote it; it just means AI could have written something similar. However, if a piece scores 40%, 50%, or higher, that’s a sign that AI played a more significant role in generating the text. I’m not comfortable with that.

For me, keeping that AI detection score under 20% aligns with my personal commitment: AI helps me structure my blog posts, but it does not write them for me. The same goes for my novels, where AI’s role is even more limited—strictly in editing and analysis, never in storytelling or prose creation.

Why This Matters

With all the conversations about AI-generated content, I think it’s important to be upfront about how AI plays a role in my process. Transparency matters, especially as readers, writers, and the publishing industry navigate the ethical and legal questions surrounding AI’s use in creative work.

For me, the bottom line is this: my books are my own. I want my readers to rest assured that when they invest their money, time and energy into reading my books, they are getting the real deal—not a cheap knockoff. When it comes to blog posts like this, I use AI to assist in structuring my ideas—but the original thoughts, the final voice, and the final message, are always mine.

So, if you ever wonder how much AI is involved in my work, the answer is simple: I may be Rambling to the Robot, but in the end, I’m always the one telling the story.

Javier

Note: The final version of this blog post generated by ChatGPT (based on the author’s dictated comments) received a Grammarly AI detection score of 15%. It took approximately 20 minutes to draft. After final edits by the author (which took about an hour), no AI text was detected.

US Copyright Office Report on Copyrightability: Copyright and Artificial Intelligence, Part 2 Copyrightability Report

© 2025 Chapelle Dorée Publishing, LLC. All rights reserved. This content may not be reproduced, distributed, or modified without permission.