Like most authors, I have been forced to reckon with AI, specifically large language models like ChatGPT, Claude, and Gemini. My journey with these tools has gone from interest, to fear, to research and development, to understanding.
Interest
Many years ago, while listening to writing blogs, I heard that AI would soon shake up the publishing world. I was interested to see how intelligent AI actually was and also a little apprehensive about how it was going to change how humans make art.
Fear
At first, I was blown away by AI tools. Computers that talk back? Cool.
I had several existential crises, afraid that LLMs were going to replace me as a writer. As a software developer for the past twenty-five years and a writer with thirty years of experience, LLMs were two worlds colliding. How could I not pay attention?
Research and Development
In this phase, I learned how LLMs could fit into my process. (Or not.) I tested their limits to see how viable the technology was for fiction writing.
I outlined a bunch of stories and tried to “write” them with Claude. The output required extensive editing to make it publish-worthy, which negated any time “saved” by having the bot generate the first draft.
Understanding
I now understand the limitations of LLMs. They don’t think, they don’t have emotions, and they can’t reason like humans do. I understand that AI-written prose will always have quirks, tics, and tells that reduce its viability as a storytelling machine for writers who know how to write themselves. I’ve also come to understand that AI is a tool, and each writer can choose how they use it or whether they use it at all.
I sometimes use LLMs for research assistance, brainstorming, proofreading, and marketing copy, which I heavily edit. I’m also not opposed to using image generators for social media and marketing.
I do not use AI to generate my prose.
During my research and development phase, I started a few stories with the help of LLMs. The generated prose required extensive rewrites, so I abandoned the process. However, a few of my published books contain very small sections of heavily edited AI-generated prose. The percentage is infinitesimally small.
Where I’m Going From Here
Erin Lucy is me. I write my prose. I write my books. And that’s the end of it.
Where We’re Headed
Some authors are all-in on AI. I follow them simply to see where the industry is going.
As for me, I find that tinkering with AI often wastes my time. I guess, when it comes to my art, I’m a luddite who likes to Keep It Simple Stupid.
I have nothing against creatives who use AI for whatever reason; however, I don’t like it when creatives don’t disclose their use, which is why I’m disclosing mine.
I have more philosophical thoughts about how LLMs might negatively affect literacy, but I’ll keep those to myself for now.
Until then, back to writing!