The AI tells have another name: cliches. If you are overusing "it's not X. It's Y" or hyphens or "The honest answer is..." I don't ultimately care if it came from AI or not, although it probably did. But I do care that the writing is boring, stale, and predictable. That was what bugged me about the Druckenmiller op-ed a few weeks ago. It was filled with cliche, poorly organized, and unreadably wonky.
Very good argument in the AI battle front. What I see mostly in the way of complaints is how it's killing creativity or it's lazy, but I see the opposite too. I see people doing clever and creative things with it, so they're not using it to put off effort but rather to extend their creative reach.
It's a bit of a witch hunt right now, the way they're going after anyone who.uses AI. it'll pass eventually, but they can do real damage to a creator for now.
I think it will eventually be like photography, where people understand that photographers, like painters, need to understand lighting, composition, etc., but understand these to be two different media.
Yes the same situation, and people also thought it would kill photography. They'll find something else to rate against eventually, but till they do keep your distance 😉
I still pick the planet over AI. Sorry, I think there’s moral obligations to its use. It kills my soul enough already knowing that I sometimes have to use it at work.
Rare, finally a sane voice in support of quality no matter which tool was used. When AI got introduced, it was as if it was permission to abdicate judgement and responsibility. And the poor reader got lost in the process. The pendulum is switching back to what good writing is and I, for one, am happy to see it.
The strongest line in these screenshots is essentially:
“Use whatever tools you want. I hold you responsible for every word you hand in.”
That gets to the actual standard: output quality and accountability, not ritual purity.
The piece also makes a useful distinction between contexts. An auto mechanic using AI to communicate more clearly is not the same situation as a student using AI on an assignment meant to demonstrate their own reasoning. That’s a much better framework than treating all AI use as morally identical.
And this part is especially strong:
“It’s the quality of the argument that’s sending it back, not the type of entity that made it.”
That’s exactly where the conversation is heading. If the prose is weak, derivative, incoherent, or psychologically false, reject it for those reasons. If it’s strong, then “you used AI” by itself doesn’t tell you what’s wrong with it.
Yay! I like to fight with AI and from that weird and wonderfully unhinged ideas happens that are quite magical - it's hard to explain, but the purity police would have my head for it
I think what concerns publishers, readers, writers, and pundits, is that anyone can just ask a chatbot to "write me a novel" and then send the result to a publisher or agent, etc. having put no effort or time into it. That is what people are upset about.
Honestly, it’s an unfounded worry spread by alarmists. I have tried it (I love trying out new technology). It doesn’t work. You can DIRECT AI to write for you, but generally only a chapter at a time. Anything more than that and it gets lost. And even then, without a lot of organisation and very specific prompting, it head hops constantly, mixes up names and plot developments, and is generally just very messy. If a writer uses AI to create their text, they are likely doing a lot of heavy editing to get it readable in a long form context.
Now, AI is getting better at it. And as more models are released, they have larger context windows which lets them remember things more clearly. And if the user is paying for a subscription, they will have access to a lot more capable models.
This is also only the case for fiction. Amazon being “full of AI books” is generally more accurate with non-fiction. Because non-fiction (barring memoirs of course, which I would lump into fiction writing) doesn’t have a set point of view or plot structures and conventions to follow, AI finds it easier to write. It’s also got a lot more data of non-fiction writing to draw from. But again, it’s far from foolproof and, unfortunately, sounds very confident when making things up. It’s a fantastic research tool, but it’s still a tool and critical thinking skills need to be employed when using it.
It’s more that what’s generated is made up of other artists works that they did not agree to have used for content generation and were not paid for that use.
"Focusing on catching people using AI misses the larger point." - agree!
Tools are just tools, and ultimately, the writer is responsible for every word in their piece. Give me a genuinely engaging and well-written point of view any day, though I'll admit I'll gladly skip even that if it contains puns (AI-written or otherwise).
It wrote of love, of grief, of pain, Of fortunes lost and found again. Its characters were rich and deep, Its endings made the readers weep.
The critics praised its splendid art, Its wicked wit, its tender heart. Until they learned, to their dismay, No human hand had shaped the clay.
They searched its pages for a flaw, Some broken rule, some fractured law. But finding none, they changed the game: The words were fine. The author, shame.
And so they banned the clever thing, While praising what its words could bring. For art, it seems, is not what you do, But whether the artist resembles you.
Completely agree. Instead of fighting this amazing AI progress, we should focus on making sure it doesn't lead to shallow, irrelevant content. Great post!
I agree with almost all of this, with one exception: I am not sure spotting AI is mechanical. I wrote an article that challenges that assumption, while also supporting your main contention that responsibility lies with the human, not the AI.
The AI tells have another name: cliches. If you are overusing "it's not X. It's Y" or hyphens or "The honest answer is..." I don't ultimately care if it came from AI or not, although it probably did. But I do care that the writing is boring, stale, and predictable. That was what bugged me about the Druckenmiller op-ed a few weeks ago. It was filled with cliche, poorly organized, and unreadably wonky.
Very good argument in the AI battle front. What I see mostly in the way of complaints is how it's killing creativity or it's lazy, but I see the opposite too. I see people doing clever and creative things with it, so they're not using it to put off effort but rather to extend their creative reach.
So intention is the starting point.
Thank you for understanding this! So many people don't.
It's a bit of a witch hunt right now, the way they're going after anyone who.uses AI. it'll pass eventually, but they can do real damage to a creator for now.
I think it will eventually be like photography, where people understand that photographers, like painters, need to understand lighting, composition, etc., but understand these to be two different media.
Yes the same situation, and people also thought it would kill photography. They'll find something else to rate against eventually, but till they do keep your distance 😉
Some authors particularly fiction authors put more energy into their anti-AI stance than their actual writing and it shows.
I still pick the planet over AI. Sorry, I think there’s moral obligations to its use. It kills my soul enough already knowing that I sometimes have to use it at work.
Rare, finally a sane voice in support of quality no matter which tool was used. When AI got introduced, it was as if it was permission to abdicate judgement and responsibility. And the poor reader got lost in the process. The pendulum is switching back to what good writing is and I, for one, am happy to see it.
The strongest line in these screenshots is essentially:
“Use whatever tools you want. I hold you responsible for every word you hand in.”
That gets to the actual standard: output quality and accountability, not ritual purity.
The piece also makes a useful distinction between contexts. An auto mechanic using AI to communicate more clearly is not the same situation as a student using AI on an assignment meant to demonstrate their own reasoning. That’s a much better framework than treating all AI use as morally identical.
And this part is especially strong:
“It’s the quality of the argument that’s sending it back, not the type of entity that made it.”
That’s exactly where the conversation is heading. If the prose is weak, derivative, incoherent, or psychologically false, reject it for those reasons. If it’s strong, then “you used AI” by itself doesn’t tell you what’s wrong with it.
Yay! I like to fight with AI and from that weird and wonderfully unhinged ideas happens that are quite magical - it's hard to explain, but the purity police would have my head for it
Can I see your AI license and registration, do you know how fast you were prompting?
Sorry officer. It's just that AI let's me create at the speed of thought, and I see how fast I can go!
I think what concerns publishers, readers, writers, and pundits, is that anyone can just ask a chatbot to "write me a novel" and then send the result to a publisher or agent, etc. having put no effort or time into it. That is what people are upset about.
And rightfully so
Have you ever tried that? It legit doesn't work.
No, have not tried that myself. But that is the worry, that AI could write novels
Amazon was said to be "full" of slop - though I do not know for sure.
Honestly, it’s an unfounded worry spread by alarmists. I have tried it (I love trying out new technology). It doesn’t work. You can DIRECT AI to write for you, but generally only a chapter at a time. Anything more than that and it gets lost. And even then, without a lot of organisation and very specific prompting, it head hops constantly, mixes up names and plot developments, and is generally just very messy. If a writer uses AI to create their text, they are likely doing a lot of heavy editing to get it readable in a long form context.
Now, AI is getting better at it. And as more models are released, they have larger context windows which lets them remember things more clearly. And if the user is paying for a subscription, they will have access to a lot more capable models.
This is also only the case for fiction. Amazon being “full of AI books” is generally more accurate with non-fiction. Because non-fiction (barring memoirs of course, which I would lump into fiction writing) doesn’t have a set point of view or plot structures and conventions to follow, AI finds it easier to write. It’s also got a lot more data of non-fiction writing to draw from. But again, it’s far from foolproof and, unfortunately, sounds very confident when making things up. It’s a fantastic research tool, but it’s still a tool and critical thinking skills need to be employed when using it.
It’s more that what’s generated is made up of other artists works that they did not agree to have used for content generation and were not paid for that use.
"Focusing on catching people using AI misses the larger point." - agree!
Tools are just tools, and ultimately, the writer is responsible for every word in their piece. Give me a genuinely engaging and well-written point of view any day, though I'll admit I'll gladly skip even that if it contains puns (AI-written or otherwise).
AI detectors are toy tech. They are also farcical. AI is being used to detect AI and getting it spectacularly wrong. Yup. Looking at you, Pangram.
ABSOLUTELY! I keep saying it: It's time to stop asking if it was AI-assisted and start asking if it's any good.
The Uninvited Author
It wrote of love, of grief, of pain, Of fortunes lost and found again. Its characters were rich and deep, Its endings made the readers weep.
The critics praised its splendid art, Its wicked wit, its tender heart. Until they learned, to their dismay, No human hand had shaped the clay.
They searched its pages for a flaw, Some broken rule, some fractured law. But finding none, they changed the game: The words were fine. The author, shame.
And so they banned the clever thing, While praising what its words could bring. For art, it seems, is not what you do, But whether the artist resembles you.
Completely agree. Instead of fighting this amazing AI progress, we should focus on making sure it doesn't lead to shallow, irrelevant content. Great post!
Oh yeah I am.
I agree with almost all of this, with one exception: I am not sure spotting AI is mechanical. I wrote an article that challenges that assumption, while also supporting your main contention that responsibility lies with the human, not the AI.
https://makingsensewithai.substack.com/p/if-this-comes-back-100-ai-i-dont?r=93w0uo