When You Edit Your Truth with AI, You Stamp Out Your Spiciness and Culture
Part 3.2 of the Book of Balance: AI Talks
I promise this is a short one today. Length consistency is not part of who I am as a writer, apparently… heh.
This is a direct sequel to yesterday’s article on AI, where I beg people to stop using every suggestion ChatGPT makes for polish and clarity, because a lot of articles on Substack these days sound like they were edited by ChatGPT, and once you notice it — as an editor like me is wont to do — you can’t un-notice it.
But you know what saddens me the most? About half of the people I see relying on ChatGPT as an editor are people whose true voices are so desperately needed.
I’m talking about neurodivergents, like you gorgeous folks with ADHD or autism (or both), or you amazing troopers from marginalized non-English cultures, particularly the ones that have been oppressed or even stamped out entirely by colonialism.
I’m gonna scream this one, so bear with me…
CHATGPT USES NEUROTYPICAL WESTERN WORDING AND YOU ARE ERASING YOURSELF BY USING ITS SUGGESTIONS!
To my fellow neurodivergent friends…
Our voices are so fucking necessary these days and I’m delighted that alla ya are starting to speak up. I need you guys, because I really don’t feel like talking about depression or ADHD as much as I want to talk about other things.
You guys speak a lot of my truths. I love reading your material. I share it a lot, because you say the things I want to say.
But when all of your articles read like a ChatGPT positivity therapy session, I admit that I start losing the message for the uncanny valley. This is because the thing that makes us who we are is our weirding way. It’s the fact that we hyperfixate on something, or go off on tangents, or talk too much (or not enough).
I’m convinced that most of you would stop the moment you realize that you are talking about unmasking, all while running your words through a machine that’s essentially masking your voice.
I want YOU! I want your realness and your messiness and your weirdness and the magical, wonderful way that you are unique. Please don’t edit yourself into homogeneity because of imposter syndrome. Your neurodivergent way of talking is exactly what we need now.
If we throw every rock we find into a tumbler, we end up with a sack of colored marbles instead of a bag of treasures, each unique with its own cracks and blemishes and quirks.
Please don’t be a smooth, polished marble. Be, instead, the cracked, rugged geode with gorgeous gems inside that I know you are.
To my marginalized and colonialized friends…
This one somehow hurts me even more, because I’m not talking to my own people. I’m talking to the people my people have oppressed.
There’s one writer on Substack that I started following recently who wrote a beautiful poem about her homeland and how it was crushed under colonialism. It brought a very literal tear to my eye.
But… every note of hers I’ve seen since sounds like it was written by a ChatGPT prompt saying, “invite readers to my Substack in a poetic, spiritual tone that evokes authenticity and softness.” Her posts, sadly, are worse.
I want to be very, very clear about something:
ChatGPT mostly scrapes Western websites and literature. It gets its philosophy and tone from Western culture. So when you use AI to polish your writing, you are destroying the voice of your people. It’s like hearing someone tell their stories about their family’s traumatic experiences with racism in the south while wearing a full-body white cloak with eye holes.
You surely get the picture.
When you’re speaking up about your ethnicity or writing beautiful poetry about your homeland, or things like that… I want to hear the real you. I don’t want to hear your story whitewashed by Western style so it’s fractionally more readable.
English isn’t your first language? So what!? I’m a professional editor and I have no interest in your perfection, so long as I hear YOU. And if all I hear is Western ChatGPT polish? I’m devastated.
You are among the most beautiful people in the world. Your ancestors probably knew more things about the truth of this world than most of us will ever understand. Please, don’t erase their voices inside you just to appeal to a, frankly, white-as-hell audience.
To reiterate a point from the last article…
Remember this quote from yesterday?
“People don't remember books; they remember sentences.”
— Morgan Housel
Please… I am literally begging… I’m so tired of hearing these polished triads or “that’s not X, it’s Y.”
Tell your story and speak your truth and when you do it, go off on a tangent that you can’t find your way back from. Use some funky phraseology that only makes sense in your original language. Those are the quirky lines that stand out and stick in our memories.
Unmask yourselves, for real. Not just for performance, but to actually speak with your own voice.
…Gosh, this style is so prevalent that I’m afraid I’m using it even when I’m not intending to (like I just did in that last line)… that’s exactly what happens when we take in things all the time that all sound the same. Even when I stopped using Chat for anything more than assessing my content for typos and repetition, I still feel like I’m mimicking its phrasing sometimes. That’s how deep this shit gets after a while.
But for reals… I need you guys. The world needs you guys. Use AI for assessing the overall content if you must. Ask if you’re being sensible. Ask if you make sense. Ask if you’re readable. Ask if you’re making your point well. Ask if you’re repeating sentiments. Ask it for sensitivity reading.
But pretty, pretty please stop invalidating your own style and perspectives by turning yourselves into generic neurotypical white voices by using all of its suggestions.
Don’t let it convince you that you need cleaning up.
I want to know the real you, mess and all.
Stay balanced, my friends ❤️🐻
Author’s Note: If you enjoyed this take on AI, consider checking out this sample of my novella series, The Vitmar Chronicles, where all of my balanced morals take the forefront in the worldbuilding! Volume II is set for release on August 22nd, 2025!
You make an interesting point I had not considered before. I usually think "too flowery" when I see raw ChatGPT output. Which seems like a femine trait... or maybe old-fashioned... I had not considered the bias in training data. Thank you.
So well said Bear. Those phrases…they are slowly infiltrating their way into all writing. I haven’t used AI because I fear them getting into my head too. Sounds a bit dramatic, I know, but we each have our own phrases and cadences…I want to keep mine untarnished!