"AI" Rejection

[AwkwardMD]

Yes, it's a different issue. I haven't done any publishing beyond Literotica, so I don't know anything about it. I don't see how there's any ethical difference between my publisher creating an AI image of two sexy people who are meant to be the characters in my novel and hiring two model/actors to portray my characters. But if there's a practical difference in terms of the reception by the book-buying audience, I would be interested in that.

One of my best friends is a graphic artist (see my art repository). From her perspective, using AI to generate graphic art is unethical in that the process rips off the talents of her and her peers BUT the use of AI to generate a story harms her not at all. What does she care if someone buys her art to put on the cover of a book that is completely prompt generated?

Let the Golden Rule guide us all and, for good measure, make friends with artists in different mediums! We're all in this together!

 

[AmonKo]

Heh, that all argument is just silly. What if you translate the story from another language to English via an online translator? It will be probably marked as AI. Everyone with a diploma knows that in school you are taught to write like a machine, AI is just doing that better. An artist has the idea but the creating process is borrowing bits you learn/see somewhere else.

 

[TheLobster]

The reason we’re “okay” with images generated by “AI” but not stories is not related to the human effort put into creating them. It’s based on how long it takes to consume them.

A 10k-word story takes like 20-30 minutes to get through, so it feels “wrong” that an LLM can produce one in 30 seconds. But an image is something you look at for a few seconds tops, and is rarely considered a work of art in its own right (e.g. book covers). Therefore it feels okay that such “utilitarian art” is automated from time to time.

If this sounds like snobbery, it’s because it is.

 

[alohadave]

The reason we’re “okay” with images generated by “AI” but not stories is not related to the human effort put into creating them. It’s based on how long it takes to consume them.

A 10k-word story takes like 20-30 minutes to get through, so it feels “wrong” that an LLM can produce one in 30 seconds. But an image is something you look at for a few seconds tops, and is rarely considered a work of art in its own right (e.g. book covers). Therefore it feels okay that such “utilitarian art” is automated from time to time.

If this sounds like snobbery, it’s because it is.

The difference is perspective. Most of us here are primarily writers, so we key in on the writing part. Visual artists are going to key in on images, and may not necessarily care about text.

Musicians are going to care more about AI generated music than text or imagery. And etc.

 

[AwkwardlySet]

The difference is perspective. Most of us here are primarily writers, so we key in on the writing part. Visual artists are going to key in on images, and may not necessarily care about text.

Musicians are going to care more about AI generated music than text or imagery. And etc.

That was also my point in a way. But it's more than just that. There are many authors here who would be happy to use an AI-generated image as a cover for their story, all the while being outraged at someone using AI to actually write such a story. It's an inconsistency or lazy thinking at best, hypocrisy at worst. Ultimately, both writing and drawing/painting are the same in this sense, regardless of the existing nuances. You can't have it both ways. Using AI for creating art is either okay or it isn't.

 

[KeithD]

I don't get this equating of the written product and its cover. The cover is packaging, not my art. My art is the writing. In the publishing world, the cover is the purview of the publisher. Often authors aren't consulted in what goes into the cover art. They certainly aren't given veto rights unless they are a huge profit generator for the publisher. I don't care if AI is used for the cover. And I don't think hypocrisy comes into that issue at all.

 

[AwkwardlySet]

I don't get this equating of the written product and its cover. The cover is packaging, not my art. My art is the writing. In the publishing world, the cover is the purview of the publisher. Often authors aren't consulted in what goes into the cover art. They certainly aren't given veto rights unless they are a huge profit generator for the publisher. I don't care if AI is used for the cover. And I don't think hypocrisy comes into that issue at all.

I'd say that it depends on whose decision it is to use an AI image for the cover. I understand that in practice it might be difficult to go to such lengths, but I was talking from the standpoint of pure principle. The reality of publishing is something else entirely and yeah, I understand that not many authors get to choose such things.

 

[KeithD]

I'd say that it depends on whose decision it is to use an AI image for the cover. I understand that in practice it might be difficult to go to such lengths, but I was talking from the standpoint of pure principle. The reality of publishing is something else entirely and yeah, I understand that not many authors get to choose such things.

The pure principle is that I am the writer not the cover artist. These are separate worlds. Even if I considered myself the artist in both worlds, one would be dominate. I don't have to have equal view of AI in use in both worlds. And I don't. I won't use AI in the writing world. I don't care if it's used for my covers. I don't think that's a bit hypcritical.

 

[storyteller92]

I won't lie...I am absolutely captivated by the passion on both sides. While I am not excited by the cracking down (I found out that my avatar was AI after stealing it from an imgur collection), I kind of get it. Lit is trying to be careful, and that is why I use multiple AI detectors in my writing and try to run them both past two editors, not just one. I also post scores and if the scores comes back as less than 90% likely written by a person, I just rewrite.
I don't know how helpful that is to the conversation, but that is my experience.

 

[AmonKo]

I won't lie...I am absolutely captivated by the passion on both sides. While I am not excited by the cracking down (I found out that my avatar was AI after stealing it from an imgur collection), I kind of get it. Lit is trying to be careful, and that is why I use multiple AI detectors in my writing and try to run them both past two editors, not just one. I also post scores and if the scores comes back as less than 90% likely written by a person, I just rewrite.
I don't know how helpful that is to the conversation, but that is my experience.

AI detectors are also AI

 

[MillieDynamite]

Short, commonly used sentences shouldn't be tagged as AI. AI does not write short sentences. It writes in almost Victorian verbosity. And yet, AI detectors will flag, "He pulled the door open." AI detectors are for shit!

 

[Portly_Penguin]

AI detectors are for shit!

And yet Lit still use it to determine whether an author uses AI and basically pushes people from their site because of it.

The irony of using a system that runs on AI to determine if someone wrote something using AI is hilarious to me. It's irony at its finest.

 

[MillieDynamite]

Actually, we don't know how or why they flag anything as AI or allow what's there to go through.

And yet Lit still use it to determine whether an author uses AI and basically pushes people from their site because of it.

The irony of using a system that runs on AI to determine if someone wrote something using AI is hilarious to me. It's irony at its finest.

 

[Mastered_again]

Short, commonly used sentences shouldn't be tagged as AI. AI does not write short sentences. It writes in almost Victorian verbosity. And yet, AI detectors will flag, "He pulled the door open." AI detectors are for shit!

I write a lot of compound sentences, 3 or 4 phrases each that some might find archaic, but then again, I learned grammar 50+ years ago, so AI detectors should leave me alone. As far as I know, no one has scanned stone tablets into their learning modules.

 

[ElectricBlue]

I write a lot of compound sentences, 3 or 4 phrases each that some might find archaic, but then again, I learned grammar 50+ years ago, so AI detectors should leave me alone. As far as I know, no one has scanned stone tablets into their learning modules.

@MediocreAuthor is working on runes, to make her diacritics legible on all devices, even mine.

 

[Portly_Penguin]

Actually, we don't know how or why they flag anything as AI or allow what's there to go through.

Not officially, no, but the rejection message implies that they are.

We've checked this work several times and it is still coming back as being composed largely of AI-generated prose.

How is it 'still coming back' as AI if Laurel is the only one reading them? If my English degree has taught me anything about reading between the lines, the wording here is the implication that an AI Checker is being used.

 

[MillieDynamite]

I should clarify my meaning. We don't know what checker is used, what the threshold for rejection or acceptance is used, or if, obviously, human writing, shitly done, is confused with AI crap. If several sentences in the light yellow on the checker might trigger a rejection, or if few dark reds might pass. We have no guide to tell what is or isn't acceptable. And from what I've read here, not much information to go by when a rejection happens.

Not officially, no, but the rejection message implies that they are.



How is it 'still coming back' as AI if Laurel is the only one reading them? If my English degree has taught me anything about reading between the lines, the wording here is the implication that an AI Checker is being used.

 

[theWollstonecraftWoman]

When I got my rejection, I just asked my editor to fix it for me. He did, problem solved.

 

[MillieDynamite]

I should further clarify. When I say shitly done, I mean those lines when you do a shity job and don't put the effort to improve it. I don't struggle with every line in the story, and I have as many shity sentences as the next person.

I should clarify my meaning. We don't know what checker is used, what the threshold for rejection or acceptance is used, or if, obviously, human writing, shitly done, is confused with AI crap. If several sentences in the light yellow on the checker might trigger a rejection, or if few dark reds might pass. We have no guide to tell what is or isn't acceptable. And from what I've read here, not much information to go by when a rejection happens.

 

[alohadave]

When I got my rejection, I just asked my editor to fix it for me. He did, problem solved.

On a side note, Mary Wollstonecraft was a clue on Jeopardy the other day. It made me think of you when I saw it.

 

[MillieDynamite]

Oh, that'll make her day. The answer, Alex, is: Who was the mother of Mary Wollstonecraft Shelly, author of Frankenstein?

On a side note, Mary Wollstonecraft was a clue on Jeopardy the other day. It made me think of you when I saw it.

 

[CatPerson]

I too am having this issue atm. One my my works is held up in this AI debacle. I posted it and after the initial rejection I , as the comments suggested, re-posted it again adding a note that said i don't use AI. It got rejected again after spending more than 10 days in pedding. Now let me start by saying I don't use any tools that have any AI like Grammarly, ProWritingAid, Quillbot or similar other to write. I do all my work on word and work alongside a Lit editor to correct any grammatical or punctuation mistakes I might have. During the week+ period my story was in pending last time i tried many of the free AI detecting tools out there to surprising results. Turns out that this AI infection runs deeper than any one in the literary community could have ever imagined.

Somehow the AI got to many of the most renowned Authors. It has broken time and space to do so. Turns out many of Tolkien works are in "fact " AI generated. So too many of Stephen King's books are full of AI generated paragraphs, some I personally tested are up too 70% AI generated. Even the bible written so long ago hasn't been spared from the onslaught of the AI.

Now in al honesty the reason this AI detection tools says absurd things like that is that they were trains on those very books, and many more of course. And AI learned to write from them. Now if someone writes in those styles the AI say that that human work is AI generated. This I find absurd, that we now have to change our own writing style just to escape the AI labeling us as AI ourselfs.

And then what? Once we all have contorted our way around the AI ver sensitive triggers, and develop brand new writing style, just so that newer AI's are now trained on those very new styles and claimed them for there own too. I for one refuse to change the way I like to write just so that some AI detection filter is happy.

I will re submit my work once more adding a note once again explaining my reasoning. But I refuse to change a single comma to it. At this point my refusal to rewrite parts of my work is more out of moral indignation than anything else. I won't allow my work or my creative process to be degraded by some mindless tool.

I understand the lit staff is between a rock and a hard place atm. AI has ,I am sure made there work so much harder, blasting them whit 100's of truly AI generated content. I haven't said at any point that AI is not a problem. But they should figure out a way to screen for real AI work and not affect actual authors here. If wait times in pending have to go up significantly to do so I don't have any problems with that. But don't ask us to change our own writing styles or intentionally make mistakes just to jump thru some arbitrarily placed hoops to make whatever program you use to check for AI happy.

 

[MillieDynamite]

For the most part, AI writes in the past perfect tense. AI uses long sentences and uses and repeats words and phrases in its works. Opulent Room, opulent mansion, opulent this, that, and the other. Therefore, if you use a particular unique word often (as in not uniquely one time), that may trigger an AI rejection. I read a work proclaiming it was written by AI, and the past perfect tense, long, long sentences, and few shorter ones drove me insane. Also, paragraphs tended to be of almost entirely the same word count or with so little variation that it became painful to read.

It isn't certain whether anyone is writing this way or not. But if you do, even if the sentences pass the AIish phrasing, the repetition of words and phrases, lack of variation in a sentence, and paragraph length might trigger some software. I'm not trying to say this is the case. I'm saying that this is something to avoid.

 

[StillStunned]

For the most part, AI writes in the past perfect tense. AI uses long sentences and uses and repeats words and phrases in its works. Opulent Room, opulent mansion, opulent this, that, and the other. Therefore, if you use a particular unique word often (as in not uniquely one time), that may trigger an AI rejection. I read a work proclaiming it was written by AI, and the past perfect tense, long, long sentences, and few shorter ones drove me insane. Also, paragraphs tended to be of almost entirely the same word count or with so little variation that it became painful to read.

It isn't certain whether anyone is writing this way or not. But if you do, even if the sentences pass the AIish phrasing, the repetition of words and phrases, lack of variation in a sentence, and paragraph length might trigger some software. I'm not trying to say this is the case. I'm saying that this is something to avoid.

Don't forget: lots and lots of coordinate clauses. Almost every sentence seems to consist of two coordinate clauses.

 

[AmonKo]

For the most part, AI writes in the past perfect tense. AI uses long sentences and uses and repeats words and phrases in its works. Opulent Room, opulent mansion, opulent this, that, and the other. Therefore, if you use a particular unique word often (as in not uniquely one time), that may trigger an AI rejection. I read a work proclaiming it was written by AI, and the past perfect tense, long, long sentences, and few shorter ones drove me insane. Also, paragraphs tended to be of almost entirely the same word count or with so little variation that it became painful to read.

It isn't certain whether anyone is writing this way or not. But if you do, even if the sentences pass the AIish phrasing, the repetition of words and phrases, lack of variation in a sentence, and paragraph length might trigger some software. I'm not trying to say this is the case. I'm saying that this is something to avoid.

I feel ya
regards
AI person