I Miss the Days Before AI Generated Comments

There was a time — I’m talking about six months ago — when a writer would post a piece of fiction online, a deeply vulnerable story about their childhood or a dying spaceship or whatever, and the comments section would roll in like a county fair. 

A man in Alaska read my story about my grandfather’s descent into dementia, and left the comment: “reminds me of the time i bought a snowmobile that leaked gas all over my front lawn. anyway good job.”

I wrote five thousand Sci-Fi novella about the loss of an intergalactic civilization, and someone responded: “This reminds me of when I had to put my dog down. And here’s a picture of my lunch.”

My autobiographical story about helping a neighbor with multiple sclerosis was read by a user mikey_d_74, who wrote a 500 word comment the length of a CVS receipt without using a single punctuation mark and explained the story reminded him of the time his sister’s Honda overheated outside of Orlando and they took it to a repair shop that overcharged.

That left me wondering if he had confused the title “An Unexpected Friendship” as the place to leave a Google review for a Jiffy Lube in Florida.

Human beings are utterly incapable of looking at a piece of art without immediately trying to find themselves in it. An author could write a historical drama set in the Napoleonic Wars and someone in the comments will invariably chime in how it reminds them of their first divorce.

Yet, the good news is that means there’s at least a living, breathing, self-absorbed creature on the other side of the internet. Even if they only read the headline, feel a tremor of something approximating emotion, get sidetracked, then decide to tell you about their life. 

Because that is what people do. People think of themselves. It is a fundamental condition of having a consciousness.

These comment leavers are self-obsessed. Off-target. Lazy. Rude. Entitled. Perhaps even narcissistic. Yet, before the age of AI, they showed up. 

They brought their whole, messy, self-absorbed selves to my front porch, tracked mud on my carpet, and left without saying goodbye. But they came.

Now? Now I publish a story and within forty five seconds, a digital entity with the processing power of a minor deity drops a five-paragraph thesis into my comment section. 

It summarizes every plot point. It notes the metaphorical trajectory of my main character's internal journey. It says it was “moved” by the story’s “emotional resonance”. It provides the most aggressively competent, average summary you’ve ever read. It reads like a book report written by a high school sophomore who is being held at gunpoint. It is flawless. 

Yet, there is no human connection in this. There is only performance of understanding. It’s like being complimented by a refrigerator. The door opens, a light comes on, and a perfectly modulated voice says, “Have a good day!”

And God help you if the AI offers “writing feedback”. 

Think about the sheer scale of AI technology for a moment. OpenAI, Google, and Anthropic have harnessed the collective knowledge of human civilization. They have fed their beasts every scrap of literature, philosophy, and code ever produced. 

And then, that AI running on a thousand servers at the Googleplex will reach into its slurry of thousands of hours of creative writing advice videos, well-meaning MFA graduates on YouTube talking about “genre conventions” and “emotional stakes,” and the largest LLM on the planet will use its cosmic intelligence to present the most generic piece of writing feedback known to man:

“You could show, and not tell, more.”

Show, don’t tell!

The AI trained on the knowledge of humanity has decided that what I need is the same advice given to a seventh-grader writing a paragraph about their summer vacation. 

Perhaps it's not wrong.

But, at least that self-obsessed man with the leaky snowmobile was wrong in an unexpected way. A person mentioning that literally any possible topic reminds them of their divorce is human. The guy who read my title about the last seagull in Sausalito and accused me of loving (or hating) a politician was operating on a wavelength I can recognize. 

So, gods of the internet, please stop the AI comments.

• Bring back the chaotic, badly formatted comment.

• Bring back the guy who misses the point so hard he lands in another zip code.

• Bring back the misspelled rage, the unsolicited advice, the stranger who says “this reminded me of me” and then vanishes into the ether like a fart in a windstorm.

At least they were trying to tell me they exist.