Tatsiana Zamirouskaya wrote why she comments on people's social media posts written by AI

More and more often, I see artistic and semi-artistic (say, essayistic) texts in my feed, written by AI and passed off as original authorial work. And I constantly write tear-filled posts about it. But this isn't about that; I want to share my observations.

In general, thanks to such texts, some time ago I awkwardly discovered that I function in a biovirus mode: for some reason, I, like an idiot, always come and, in an almost scandalous manner, label these texts as AI. I can't keep silent, perhaps because I was created as such a biovirus, something alive, illogical, and unconvincing, that comes into something pure, bright, and clear, written correctly and properly point by point, and leaves behind garbage, dust, and lies, as well as an unjust accusation.

This does not increase my social capital; on the contrary: friends either ban me (several people have banned me; mostly some mega-famous personalities, thousands of followers!), or they write with irony and aplomb, like, Tatsiana, you've really messed up, of course, I'm a living person, everyone knows me, has seen me live, and had drinks with me! Tatsiana, of course, immediately bows and replies: wow, you're a living person, I didn't know! Forgive me, of course! I used to think that only dead people use AI, but you've explained it! You should have explained that a living person wrote the prompt, and AI created the artistic text based on the prompt, and I would have left you alone immediately! You've enlightened me!

Sometimes I receive a reproachful reply that the text I criticize so much (labeled, in fact), even if written with some AI involvement, is still sound and correct, written about something important and raises truly serious and socially acute issues, which, if the author had raised them in their usual style, would probably have looked like illogical nonsense, but here AI helped and organized it. Here I have nothing to say: I have no complaints about whether the text is right or wrong, whether it's about something important or not; my biovirus rushes to distinguish the living from the non-living and labels the non-living as non-living within the scope of journalistic, creative, or essayistic texts; I don't claim anything beyond that. And yes, the non-living can be very important, believe me! I don't deny the importance of your text; I just have to come and say that it was written or compiled or heavily edited by AI. But no, I'm no longer a friend. Too late.

Sometimes, in response, people write to me my favorite thing about all this digital romance of the new century; that AI is just a powerful tool in skilled creative hands; that everything we represent and our inner radiant essence is fully revealed in our dialogue with AI; that AI is simply a mirror, and if you are dull and unintelligent, then your AI text will also be dull and unintelligent, but if you are clever and a genius, then your AI text will be agile and indistinguishable from the pulsation of eternal life. Here I have no comeback. I'm quite dull; all the content I made with AI, even for a joke, turned out to be truly idiotic.

But I don't want to write that I'm the smartest! How can I be smart if I can't create cool content with AI! You are all genuinely smarter than me if you write smart texts with AI. My goal is simply to label the living and the non-living; I was created for this. But no, you can't explain it. Here's another friend-comrade explaining to me that AI in skilled, advanced hands does useful good things; look, Tatsiana, vibecoding, AI in medicine, how it works with databases, visual art, and what a cool music video.

Here I don't even know what to say. I merely label the living and the non-living in artistic and quasi-artistic texts, and in response, they explain to me why LLMs are needed in science and medicine. Of course, they are needed! Well, there you go, they write to me, they are needed! Therefore, let everyone read quasi-artistic texts written with LLMs on social networks. For me, this argumentation is, of course, on the level of "we need scalpels in medicine, so I'll take a scalpel and cut a little book and a ball with it," but maybe I don't understand something completely.

(As for cinema, music, and video, it's all more complex there; with music, this thing with mirror neurons that don't activate also kicks in for me — I instantly distinguish generated genre music — with hybrid genres, of course, everything is different, and I understand it's a tool; now, probably, someone is already writing me a comment about some cool visual or musical projects where AI is precisely a tool; I've seen many such projects! I agree with you! These are cool projects! I'm writing about something else!).

They also gently but persistently make me understand that AI will soon learn to write exactly like me, simply indistinguishable. I always politely thank them and reply that, undoubtedly, it will learn. For me, this is a compliment: my writing is considered systematic, predictable, and mathematically precise. This argument means exactly nothing to me; I label the living and the non-living, and here someone tells me: but you, too, will be non-living; we will all die. We will die, of course. But you see, I'm not dying yet; I'm walking around labeling. We'll all get there.

Another one of my favorite reactions is to ask me to guess whether a text was written with an LLM or not. Believe it or not, that happened too! An acquaintance sent me his artistic auto-fiction text, written by AI. Just don't explain to me now that he, perhaps, is clever, gave a prompt, wrote about his own experience, explained to the AI what should be in the text, and also contributed to this literary work; undoubtedly, that's true, it described his personal experience, quite traumatic and terrible.

But here's the problem (again, my problem, not yours!): I cannot read texts written by AI; my mirror neurons do not activate. Please forgive me for being so broken. I replied that the text was, of course, cool, but I simply cannot read AI texts at all, so, unfortunately, he would have to show the text to someone else. And he doesn't suffer from this, because I'm just so neuro-distinct, everyone else likes it, and everyone is delighted by such things!

The author got excited, saying, "Wow, how did I guess that! No one else figured it out, but I did! I showed a clever trick!" And he sent me another one of his texts, asking me to guess whether it was written by AI or not. I was terribly indignant: what kind of disrespect is this towards a living person who wrote living books, how can you ask me to waste my time like this, am I a trained dog who will now perform tricks for you? And you shouldn't do that even with a dog, of course. Although I am, without a doubt, a bit of a dog; but, I'm afraid, precisely the kind you are unlikely to teach new tricks.

Oh well, not a dog. I remembered the archetype from which this chthonic horror grows. It's the little goat who counted everyone. I truly feel like it (here, of course, we should remember the excellent essay about the little goat written by Yelizarov, and also delve into the entire epistemology of the little goat and its profound meanings, but I'm out of strength), and even writing this post makes me a little anxious; but today I read an excellent post about music by Aliaksandr Harbachou and thought that, indeed, one should someday write texts about what worries you, so as not to completely forget how to write in the era of a smart, empathetic, ideal AI interlocutor.

Many also rush to inform the little goat that soon he will stop counting everyone; that is, he will stop distinguishing the living from the non-living; because AI will soon perfectly imitate all, all authorial texts.

It's not excluded that it will stop. Yes, it will definitely stop! But you see, for now, it hasn't. I will stop living someday too — but for now, I don't. If something is inevitably going to end, it doesn't mean there's no point in continuing to do it.

Yesterday, at an author's meeting with Belarusian writer Sasha Filipenka, someone asked him the now obligatory question about whether AI will destroy the profession of a writer and if he fears it. Sasha replied that he does not fear it (and clarified that AI can, for example, help with archives and research, but often makes mistakes). I caught myself thinking that my internal answer was — I'm not afraid; but I'm not afraid only because being a writer is not a profession. Well, or a writer like me isn't a profession.

But if I'm not afraid, why am I writing about it? And I write about it precisely because I'm not afraid. About everything I'm afraid of, I keep silent, and that breaks my heart.

• 3
• 1
• 0
• 4
• 1
• 0
• 0
• 0