This is so well thought out, researched, and presented. It is thoughtful and logical and kind. I agree with you completely and WISH more people would just.. be open and curious enough to take the time to LEARN and understand what it means to use AI in writing. It is a spectrum, not a definitive static point. For what is worth I've re-stacked it. Thank you for taking the time to write this.
Thank you so much for the restack and very generous reading ❤️ The sad joke underneath all this is that the "detection crowd" outsourced the reading half instead of the writing half. And curiosity... curiosity has left the chat.
I am a cinematic writer, so AI helps to fill in the in-between for me. Even if I go back to read it and rewrite, because, let's face it, AI makes mistakes- we all do. If it's not in my voice, I scrap it and restart, fixing words. Then you bring it to Substack or anywhere to publish your chapters, and you are curious what the AI detector says, and it's 100% AI. Excuse me, did AI make my Aria Valerius or Jason Mendoza? No, I did. I ran my work through a recent chapter, and "As I" was the only words that were not AI-generated. I worried over so many things. But thank you for this article bit of information that sits inside my chest at the end of the day, wondering if I should just throw away everything and never let my world out.
I think the biggest risk is that you start writing to fool a classifier instead of to create something interesting for a reader. Don't throw the world away over a number that has never once been moved by a story ❤️
ya I almost threw my first book into the trash on my computer 7 times because of the witch hunt going on with AI detectors. It took a lot to just keep focus cause I wanted my world out there.
I genuinely look forward to receiving your “new posting” notification in my inbox because your writing is always brilliant and, particularly in today’s case, so razor-sharp.
I too had a recent, insulting AI-adjacent experience that echoes your thoughts: As a longtime editor, things like m-dashes and their ilk are like breathing to me, and I had no idea that apparently they are considered a dead giveaway for AI usage in writing. So I was shocked when someone recently accused me of using AI in a comment I’d written on TikTok (I wouldn’t even know how to, and anyway, why would anyone bother in that setting and scope?). It was insulting and wild that I had to consider defending my learned abilities to someone who couldn’t conceive that perhaps I actually knew how to use those punctuation marks all by myself. I cannot imagine having to wade through that frustration and insult every day just to be able to publish your own thoughts.
Oh, thank you so much, that means an incredible amount to me! Especially since I was always the person who struggled with writing essays at school :)
Your TikTok em-dash story is exactly it. When ChatGPT first appeared, I was amazed by how many people didn’t know the difference between a hyphen and a dash, or that pressing the hyphen key twice on your phone automatically turns it into a dash. The same goes for different types of quotation marks and so on. Maybe I’m just a perfectionist with OCD tendencies, but I was genuinely surprised.
And the most frustrating part is that this witch hunt, conducted with the help of AI, which makes it especially ironic, is worse than the use of AI itself. Accusing someone of writing with Claude costs nothing, but defending yourself sometimes requires a 3,500-word essay...
I do have one slight push back, if I may 😊 I do not think that all forms of human collaboration, mixed processes etc. with people involved at various stages etc. are ethically comparable to AI-assisted writing (or any creative output, for that matter), simply because we cannot trace back where the idea and words are from in the same ways we can name a collaborator. Of course, ideas get shared imperfectly, we do not always remember who said what, attribution is not straightforward — but I still think it’s not directing comparable to AI. Thinking along with someone else also may be a waste of human resources (time, money), but not with at the same scale and with the same implications as AI. What do you think?
Anyways, thanks so much for this great text. I’m happy I stumbled upon it 😊
Wooooow, I read your piece practically clapping my hands! Did I really stumble into a fairly academic framework all by myself, simply by thinking about perfume? It really is true: “if only you knew from what rubbish / Poetry grows, knowing no shame…” https://ruverses.com/anna-akhmatova/i-have-no-use-for-regimental-odes/12229/
As for the pushback — yes, I expected someone to raise this. I started with the comparison between synthetics and naturals in perfumery, but very quickly wandered into a broader argument about authorship.
I think that, over time, AI will become a completely legitimate figure in the creative process: a co-author, editor, ghostwriter, assistant… much as it is already becoming for software engineers. It can help them brainstorm an architecture, write the code, cover it with tests, and review it afterward. And there is artistry involved there too, even if perhaps less of it. So I think that, eventually, we will have to define this category as well.
Thank you again 💖 This is really the best response I could have hoped for.
At the risk of becoming the boring theorist in the room, what you’re referring to is—having ideas that echo someone else’s without fully knowing it, a little bit like ideas flowing all around—has been theorized with the concept of ‘polyphony’ introduced by Bakthin. If you’ve never stumbled upon it, you may find it interesting.
It really repels me when everything is nowadays about proving yourself not being AI. So much time wasted because in the first place the AI slop shouldn’t be provided - rather using it for corrections, data finding etc than “ creative “ work.
I would go even further: I sincerely don’t see anything inherently wrong with AI generating the prose, as long as the thoughts, taste and judgment behind it belong to the person publishing it. If the result is interesting and well written, I’ll read it. LLMs are, let’s admit, very good at producing competent prose. What they cannot produce on their own is meaning. And, again, let’s admit, humans were perfectly capable of manufacturing slop at industrial scale long before LLMs — a good example of this is even our beloved Instagram perfume content.
Thank you for this beautiful essay. While I believe that much of the current hysteria around AI comes from a fear that those of us who wrote "the old fashioned way" suffered for nothing, much of it is also around the unravelling of the simplistic idea of authorship we hold. Of course this is not new--Barthes' essay was published in 1967--but this is the newest iteration of it. I suspect that over time we will come to look at AI assistance much like we look at photography.
This is so well thought out, researched, and presented. It is thoughtful and logical and kind. I agree with you completely and WISH more people would just.. be open and curious enough to take the time to LEARN and understand what it means to use AI in writing. It is a spectrum, not a definitive static point. For what is worth I've re-stacked it. Thank you for taking the time to write this.
Thank you so much for the restack and very generous reading ❤️ The sad joke underneath all this is that the "detection crowd" outsourced the reading half instead of the writing half. And curiosity... curiosity has left the chat.
I am a cinematic writer, so AI helps to fill in the in-between for me. Even if I go back to read it and rewrite, because, let's face it, AI makes mistakes- we all do. If it's not in my voice, I scrap it and restart, fixing words. Then you bring it to Substack or anywhere to publish your chapters, and you are curious what the AI detector says, and it's 100% AI. Excuse me, did AI make my Aria Valerius or Jason Mendoza? No, I did. I ran my work through a recent chapter, and "As I" was the only words that were not AI-generated. I worried over so many things. But thank you for this article bit of information that sits inside my chest at the end of the day, wondering if I should just throw away everything and never let my world out.
I think the biggest risk is that you start writing to fool a classifier instead of to create something interesting for a reader. Don't throw the world away over a number that has never once been moved by a story ❤️
ya I almost threw my first book into the trash on my computer 7 times because of the witch hunt going on with AI detectors. It took a lot to just keep focus cause I wanted my world out there.
I genuinely look forward to receiving your “new posting” notification in my inbox because your writing is always brilliant and, particularly in today’s case, so razor-sharp.
I too had a recent, insulting AI-adjacent experience that echoes your thoughts: As a longtime editor, things like m-dashes and their ilk are like breathing to me, and I had no idea that apparently they are considered a dead giveaway for AI usage in writing. So I was shocked when someone recently accused me of using AI in a comment I’d written on TikTok (I wouldn’t even know how to, and anyway, why would anyone bother in that setting and scope?). It was insulting and wild that I had to consider defending my learned abilities to someone who couldn’t conceive that perhaps I actually knew how to use those punctuation marks all by myself. I cannot imagine having to wade through that frustration and insult every day just to be able to publish your own thoughts.
Oh, thank you so much, that means an incredible amount to me! Especially since I was always the person who struggled with writing essays at school :)
Your TikTok em-dash story is exactly it. When ChatGPT first appeared, I was amazed by how many people didn’t know the difference between a hyphen and a dash, or that pressing the hyphen key twice on your phone automatically turns it into a dash. The same goes for different types of quotation marks and so on. Maybe I’m just a perfectionist with OCD tendencies, but I was genuinely surprised.
And the most frustrating part is that this witch hunt, conducted with the help of AI, which makes it especially ironic, is worse than the use of AI itself. Accusing someone of writing with Claude costs nothing, but defending yourself sometimes requires a 3,500-word essay...
We live in interesting times.
Brilliant! So much food for thought.
Thank you so much, Nina. It was really hard to write it for many reasons, so I'm very glad it resonates.
Thanks so much for this fascinating read. I’m sharing a text of mine which shares most of its premises, in which what you describe as authorship (“the signature beneath it”) is conceptualized as *principalship* by Goffman: https://languageunbothered.substack.com/p/what-goffman-would-say-about-ai-assisted?r=eyul3&utm_medium=ios
I do have one slight push back, if I may 😊 I do not think that all forms of human collaboration, mixed processes etc. with people involved at various stages etc. are ethically comparable to AI-assisted writing (or any creative output, for that matter), simply because we cannot trace back where the idea and words are from in the same ways we can name a collaborator. Of course, ideas get shared imperfectly, we do not always remember who said what, attribution is not straightforward — but I still think it’s not directing comparable to AI. Thinking along with someone else also may be a waste of human resources (time, money), but not with at the same scale and with the same implications as AI. What do you think?
Anyways, thanks so much for this great text. I’m happy I stumbled upon it 😊
Wooooow, I read your piece practically clapping my hands! Did I really stumble into a fairly academic framework all by myself, simply by thinking about perfume? It really is true: “if only you knew from what rubbish / Poetry grows, knowing no shame…” https://ruverses.com/anna-akhmatova/i-have-no-use-for-regimental-odes/12229/
As for the pushback — yes, I expected someone to raise this. I started with the comparison between synthetics and naturals in perfumery, but very quickly wandered into a broader argument about authorship.
I think that, over time, AI will become a completely legitimate figure in the creative process: a co-author, editor, ghostwriter, assistant… much as it is already becoming for software engineers. It can help them brainstorm an architecture, write the code, cover it with tests, and review it afterward. And there is artistry involved there too, even if perhaps less of it. So I think that, eventually, we will have to define this category as well.
Thank you again 💖 This is really the best response I could have hoped for.
At the risk of becoming the boring theorist in the room, what you’re referring to is—having ideas that echo someone else’s without fully knowing it, a little bit like ideas flowing all around—has been theorized with the concept of ‘polyphony’ introduced by Bakthin. If you’ve never stumbled upon it, you may find it interesting.
Have a wonderful evening ✨
This was a very interesting read!
It really repels me when everything is nowadays about proving yourself not being AI. So much time wasted because in the first place the AI slop shouldn’t be provided - rather using it for corrections, data finding etc than “ creative “ work.
Great tool has become almost an enemy.
I would go even further: I sincerely don’t see anything inherently wrong with AI generating the prose, as long as the thoughts, taste and judgment behind it belong to the person publishing it. If the result is interesting and well written, I’ll read it. LLMs are, let’s admit, very good at producing competent prose. What they cannot produce on their own is meaning. And, again, let’s admit, humans were perfectly capable of manufacturing slop at industrial scale long before LLMs — a good example of this is even our beloved Instagram perfume content.
Yes, agree with the human slop too :)
What a brilliant analogy! 👏 I see it the same way!
Thank you for your kind words. I'm really glad that you enjoyed it!
Thank you for this beautiful essay. While I believe that much of the current hysteria around AI comes from a fear that those of us who wrote "the old fashioned way" suffered for nothing, much of it is also around the unravelling of the simplistic idea of authorship we hold. Of course this is not new--Barthes' essay was published in 1967--but this is the newest iteration of it. I suspect that over time we will come to look at AI assistance much like we look at photography.
Since you’re mentioning Barthes, I couldn’t help but share this essay of mine (my first on Substack!) in case it resonates: https://languageunbothered.substack.com/p/death-of-the-author-rebooted I hope it’s okay, otherwise I’ll of course delete my comment 😊