The Substack Writer's Journey With AI
Not Just Another AI Article
You found your voice. Maybe it happened in a moleskin notebook or during the years you sculpted away at your plots with a chisel to your heart and pen in your fingers. Perhaps it snuck up on you between the sheets stained crimson or when you smelled her summer’s lavender. Or did it come to you when the ring on his finger pressed too pleasantly against your throat? Whatever it was, you found a way to see through yourself and into others. You’re ready to publish on Substack.
No, you don’t try to sound like you’re talking to your best friend. You sound like an idiot with your best friend, especially to others without context. Instead, you’ve cultivated a syntax where 10 words take five minutes to string together. It’s yours (you just gotta throw in some personal asides), give an immediacy that makes your wit seem extemporaneous, and seamlessly show your academic vocabulary while making it clear you’re no stranger to the stresses of the streets.
People don’t read past the second paragraph. The ‘wants’ aren’t clear, the stakes aren’t high enough, and it doesn’t resonate. Just analyze the open rates, likes, and comments—see what sticks.
That’s what the successful Substackers substacking about success on Substack tell you to do. They also tell you to use more white space. You should also bold what’s important.
Make Sure to Use Headlines.
Some people are here to skim.
So you do that.
I mean, why not?
You were always into switching it up, and come to think of it, reading a paragraph from 18th-century literature on your phone is a pain in the ass.
You kill your darlings. And remember that 800-1’200 words is the sweet spot. Perfect. You don’t have time to write more than that anyway. It’s essential to be consistent and not worry too much about perfection. You gotta form a community. Comment everywhere, publish several Notes a day, and publish at least once a week. Three if you can.
Substack isn’t paying the bills, though. You have another job. It’s also in front of the computer. You publish Notes while you eat, comment while you shit, and restack when a commercial comes on because you down-graded to poor people Netflix. Now you have more money to subscribe to more Substackers.
Humour should be relatable. It should also follow the benign-violation theory.
Keep that in mind.
Oh, and the white space.
Anyway, you can’t find time anymore. You want to leave long, thoughtful comments, but you have to get back to the job that pays you. Are you really going to set aside two hours at the end of the day to reply and comment thoughtfully?
Yes.
Well, no. Life gets in the way. Plus, you just saw a Note telling you to get out in nature. You really want to because the anxiety in your spine and cotton-dry eyes are begging you to find peace among the silent pines. You rush.
Here Comes AI
You can try ChatGPT, DeepSeek, or whatever else. You quickly jot down your ideas and ask it to add some references you don’t have time to look up.
The person responds to your AI comment thoughtfully, without showing off cited knowledge, and with genuine gratitude. Shame creeps up. Who the hell are you? You’re a brand. At least, that’s what the mentors told you when you couldn’t figure out why nobody was reading your work.
You blame life. Your circle is getting smaller—only the best remain. The best, it turns out, don’t even read your work. You think about the person who left you. In hindsight, that ring was always a symbol of their betrayal and your masochism. Now, lavender tastes like soap.
⬆ Was that paragraph any good? You’ve been rejected by publishers for years, manipulated by gurus, betrayed by lovers, and supportive voices do little to keep self-doubt from weighing you down. Ugh, so much for having a fresh and unique voice.
You ask ChatGPT to make it sound more like the writers you idolize, like Jack Kerouac.
You’re getting older, man, the years slipping past like rain on a bus window, and the circle tightens—only the real ones stick, the ones who don’t need to read your damn words to know you, to know what’s been rattling in your bones all along. And that ring—Christ, that ring—always a bad omen, a little noose on your finger, a joke the universe played at your expense. Now lavender ain’t romance, ain’t nostalgia, just the scent of gas station soap and hands scrubbed raw.
And was that any good? Hell if you know. The rejections stack up like old newspapers in a shut-in’s house, the gurus with their snake oil pat your back, tell you to "believe," but belief don’t put ink on the page or keep the doubt from slithering in at night. Lovers, mentors, all ghosts now, and the "supportive voices" sound tinny, far away, like a radio tuned to a station that don’t quite reach your part of the world. Fresh and unique voice? Ha. Might as well be shouting into the void.
What the fuck is this? It isn’t Kerouac. And it sure as hell isn’t you.
How? ChatGPT already knows everything about you because you ask it every question that comes to your mind. Google takes too long. You even use ChatGPT as a therapist because it’s less expensive than weekly sessions and less intimidating than the suicide hotline.
You feed it various prompts and ask it to analyze dozens of your texts to get a sense of your voice.
Was that paragraph any good? Who knows. Publishers don’t think so. Gurus would tell you to manifest a different outcome. Lovers promised they believed in you, right up until they didn’t. And the supportive voices? They’re nice, sure, but they don’t stop the doubt from curling up in your chest like a cat that refuses to be moved.
So, you ask ChatGPT to make it sound more like the writers you idolize, like Kerouac—because if you can’t be fresh, at least you can be borrowed.
Huh. Does that sound like you? To be honest, you don’t know who you are anymore. An empire is collapsing, the world’s on fire, and since you’re a sensitive artist, you feel all the pain.
Yet, the more you use AI, the less you feel. You ask it to give you advice—it’s no different than the guru who charged you $150 for a one-hour session.
You publish your first piece with AI. It gets just as many likes—or lack thereof—as your other pieces. It affects you less, though. It didn’t take as long to write. Plus, it’s not you they rejected.
After reading over your article with ChatGPT as your editor, you realize something is off. Instead of relying on your gut, you turn to an AI detector. Turns out, “Lovers promised they believed in you, right up until they didn’t,” and “They’re nice, sure, but they don’t stop the doubt from curling up in your chest like a cat that refuses to move” is likely AI-generated.
You learn about tokens and how LLMs generate human-like responses. They have nothing to do with literature or craft. They don’t even fully align with morphemes, phonemes, or the linguistic structures you studied. You’re lost and about to click “Humanise text,” but your last bit of artistic dignity screams, “This has gone too far.”
You realize that AI isn’t an editor. It’s a thief who wants to turn your voice into data and humanity into productivity. Then, you take a look through your moleskin notebook and, once again, dive into the mistakes that left your sheets stained crimson and a ring pleasantly pressed against your neck. You play.
Born Without Borders is a reader-supported guide to building bridges across divides, cultural psychology, and how to salir de las fronteras que impone tu mente. If you want to support my work, the best way is to take out a paid subscription for $5/month or $30/year. You can also Buy Me a Coffee.
If you enjoy my work, please consider sharing it with others.
More by Born Without Borders
Love it! This comment is 💯 not AI generated!
Since four months ago that i've started learning German, i've used AI (Microsoft Copilot) more frequently than before, but only when i don't have much time for doing my homeworks. About usig AI, I'm in league with those who believe that humans must control AI, not the other way around.