We Know AI Wrote Your Comment
What’s the point
My Medium friends can read this story over there as well.
Scroll through any Medium story or Substack post these days, and we all see them.
The comments:
“Such a thought-provoking insight.”
“Thank you for sharing this valuable perspective.”
“This really made me reflect on the future of creativity.”
Robots are writing the articles. And now they are also pretending to be the readers.
I keep publishing stuff on Medium and Substack and underneath come these little paragraphs of “nothing”. All the time!
Not disagreeing. Not adding. Not asking…
What’s the point?
Comments
I love the genuine comments I get. People asking things, telling me cool anecdotes from their point of view on the topic, or something like that.
I wish I could read more of those. But I get a ton of comments now, on multiple platforms. I can’t keep up with the responses. And unfortunately, that’s also due to the fact that half of the comments now are so blatantly AI-generated.
Comments used to be messy, sometimes funny or irritating, and full of typos. The good old days. That’s how I still write my comments.
They were human. Imperfection is now a premium asset.
Now, half the replies are generated because someone read a growth hack on X that said “engage with creators to grow your personal brand” and then automated the whole thing.
It’s noise
The defenders of this stuff will say, hey, engagement is engagement.
The algorithm loves it. More comments, more reach. Might be true…
But this kind of engagement is just bad.
It trains everyone to stop caring. When we scroll through ten fake comments, we stop reading them at all.
Then the one real person who actually took time to write something thoughtful gets buried in a pile of AI garbage.
And to be fair, it’s not even AI that’s the problem here. AI could write helpful comments if the user wanted it to. But they don’t care what’s actually in it. Just publish quickly.
Why?
Why people do it
I think people use AI comments for two reasons.
One, they want attention. They hope the author clicks their profile, sees their Substack, and subscribes. I don’t think that works out too often, with this comment quality.
Two, they’re lazy. They want to look supportive without actually thinking. Writing a real sentence feels too hard. So they outsource their personality to ChatGPT.
But again, they do it in the worst way. ChatGPT would be capable of writing a good comment. Instead, they just let it write garbage because they can’t even write a simple yet proper prompt.
It’s oddly disrespectful
Writers can tell. Humans can tell.
We don’t need AI detectors. If there ain’t one typo in it, it’s probably AI.
Just kidding. But you know what I mean. We usually detect AI pretty quickly.
Real people ramble. They misunderstand things. They get emotional. AI comments don’t.
The Bottom Line
You know… I get it. Reading takes time. Commenting takes time, too.
So, if we don’t have the time, let’s just not do it. But if we invest the time and want to comment, let’s do it right. Even if we use AI, we should use it to generate a proper comment with a cool thought, a question, or something helpful.
That will help engagement. And SEO. And more. It might also get the author to actually look at it and check out our profiles.
A comment does not need to be smart. It does not need to be long. It just needs to be half-human and useful.
P.S.: To everyone who has written great, helpful, and critical comments on my stories: thank you so much for taking the time! Second, I am sorry if I haven’t gotten around to responding yet. I try to read as many as I can and respond to questions.
