Transparency note: As of January 2026, I’ve deleted my ChatGPT account and no longer use it in my writing processes. Even as I wrote this blog post a few months prior, I was flirting with the idea of getting rid of it altogether. While I stand by the approach outlined below to use it thoughtfully and minimally, I just don’t want to outsource my creativity and thinking to AI at any level anymore, at least for now. I don’t think everyone should do the same—do you. I stand by the advice here as the best way to use AI in your writing, should you choose to do so.
This is an excerpt from a longer post I published on Substack—most of which is drawing comparisons between Selling Sunset and ChatGPT, so if you’re into random musings like that, read it HERE.
The below bit is most relevant to my work as a freelance content writer. If you are considering working with me and have questions or want to talk more about how I use AI, I’d love to have that conversation. Reach out at any time.
Because AI is everywhere and there are conflicting opinions on what the “appropriate” use of AI is as a writer (or insert profession of choice here), I think it’s important to provide some clarity on how I personally think about AI and use it in my work.
I believe this is important for transparency reasons, so my clients know what they’ll get from me, what my Substack readers know they’ll get from me. And I think it’s also important for accountability reasons, so I’ll keep remembering that AI makes me feel dumb and, thus, need to limit my use of it.
While I don’t want to lock myself into a never-changing and rigid set of rules, I’ll take this opportunity to outline my broad concerns with using AI as a writer and my response—my personal AI policy.
My concerns with using AI as a writer:
- Using AI to write makes me feel dumb because I see how it truncates my creative process. As this article calls out, creativity is a use-it-or-lose-it muscle to exercise, and we are (I am) quickly losing it if we (I) rely on AI to write and generate ideas.
- AI is not really that good yet, so I can’t create high-quality work without heavily supervising it and editing its output. AI “slop” is a real thing (also sometimes funny), and I don’t want my writing to be like that.
- LLMs like ChatGPT were trained on the work of other writers (which is why the em dash debate is so annoying—writers used it first, AI copied it, and now everyone thinks it’s a sign of AI). While there is some nuance to this discussion, this is basically plagiarism. Here’s a good take on this idea.
These concerns are in addition to general concerns about AI’s negative impact on the environment, job market, and our collective ability to have genuine, human relationships (this article does a good job exploring the nuances of human-ai relationships).
That all said, AI is here to stay. And I don’t think it’s possible to fully rid yourself of it as a person who spends time on the Internet.
I agree with the general sentiment that AI is a tool, and you can use it the way you want to. For me, I want to use this tool minimally and carefully—not often, not as my go-to, and always with the mindset that this tool can quickly overpower my own brain power and effort.
Some ground rules:
- AI comes second. I must first spend time thinking, writing, outlining, and researching on my own before I open the app.
- AI needs to be trained and directed. I tell ChatGPT what I want in terms of tone, style, content, etc.—I don’t let it generate without my direction.
- AI needs to be heavily edited. I get last say over what is produced. I need to look with my human eyes and brain, editing the heck out of it. Human editing is necessary for accuracy, yes, but also style, tone, and intended meaning.
Like I said, I’m not trying to be black-and-white about using AI, but I do want to be mindful about how I use it. I’m still trying to figure out how (or if) to use it in my work. All I know now is that I don’t want to outsource my creativity and our collective humanity to LLMs that were trained to think and talk like us… They’re not us. We, the people, have the power! I want to read things written by other humans, and I want people to read what I write and connect to it on that human level.
Part of my purpose of publishing this is to get some clarity on how I am thinking about AI today, in November 2025, so I can keep revisiting it and updating it over time. I think it’s important to be flexible and open-minded in this convo and understand that things change, we change… And that’s okay.
Leave a Reply