The AI Rejection Conversation Matters

[Voboy]

AI is a problem. That problem is likely to get worse.

NO institution has figured out how to deal with that problem. Many of those institutions have vastly more power, money, time, and staff than Literotica does.

Ergo, expecting Literotica to "solve" AI is, in my view, naive. I think there will be many, many, many growing pains involved as we stupid-ass humans surrender so much control to these machines we're enabling to replace us.

I don't believe, with some optimists, that this "AI rejection debate" is a temporary problem with a solution likely to manifest itself soon. Rather, I think this is the new normal. It might be wise to start thinking about how to accommodate ourselves to that new normal.

Or? Maybe I'm wrong. Either way, I think expecting this site to "get it right" is expecting too much, and I mean no offense to Laurel and Manu when I say that.

 

[EmilyMiller]

Yeah, we could, but right now Literotica is basically putting a band aid on a larger gaping wound if we just focus on the now, it doesn’t heal it, it just relieves the bleeding. But it will get infected. The wound will get worse I’m afraid. Just worrying on the now is a big mistake as it concerns the future of not just this website but any website that lets people publish their own stories like Wattpad. It’s a big problem that’s not going to get better.

In your opinion. Based on your speculation.

Emily

 

[EroticPower]

AI is a problem. That problem is likely to get worse.

NO institution has figured out how to deal with that problem. Many of those institutions have vastly more power, money, time, and staff than Literotica does.

Ergo, expecting Literotica to "solve" AI is, in my view, naive. I think there will be many, many, many growing pains involved as we stupid-ass humans surrender so much control to these machines we're enabling to replace us.

I don't believe, with some optimists, that this "AI rejection debate" is a temporary problem with a solution likely to manifest itself soon. Rather, I think this is the new normal. It might be wise to start thinking about how to accommodate ourselves to that new normal.

Or? Maybe I'm wrong. Either way, I think expecting this site to "get it right" is expecting too much, and I mean no offense to Laurel and Manu when I say that.

I agree, Literotica can’t solve this problem of AI. But I feel like transparency might be a temporary solution, I believe writers if they have used AI to write have to tell their readers like a disclaimer that AI was used in the creation of their work and they have to mention how much of the AI they used for it (like was it the whole thing, to help with some spelling etc) If a writer doesn’t tell their readers that AI was used, then they risk having their work rejected. Obviously it doesn’t help the problem with people’s work who doesn’t use AI getting rejected though but it’s a start. There’s no perfect solution to this. This is just a solution that someone on the forums recommended and while it is is a bad idea but there isn’t an idea that is a good one at the moment, just found the one that was less s**t.

I know other websites have just basically given up on AI moderation and just letting stories in. It makes me curious about their views on AI, like why aren’t they doing the same things like Laurel is doing especially when things could change in the future about AI and how much it could land those websites in trouble.

 

[Voboy]

I agree, Literotica can’t solve this problem of AI. But I feel like transparency might be a temporary solution, I believe writers if they have used AI to write have to tell their readers like a disclaimer that AI was used in the creation of their work and they have to mention how much of the AI they used for it (like was it the whole thing, to help with some spelling etc) If a writer doesn’t tell their readers that AI was used, then they risk having their work rejected. Obviously it doesn’t help the problem with people’s work who doesn’t use AI getting rejected though but it’s a start. There’s no perfect solution to this. This is just a solution that someone on the forums recommended and while it is is a bad idea but there isn’t an idea that is a good one at the moment, just found the one that was less s**t.

I know other websites have just basically given up on AI moderation and just letting stories in. It makes me curious about their views on AI, like why aren’t they doing the same things like Laurel is doing especially when things could change in the future about AI and how much it could land those websites in trouble.

Honestly, I'm not sure it matters all that much.

AI produces shit stories. So do some human writers. This site (and many others) proves there's a market for shit stories, regardless of the navel-gazing of those of us in the AH (who genuinely do care about excellence in writing, I think). We're a tiny minority in a sea of preferences.

So it's not that clear to me how "dangerous" AI content is to those of us who plug away at this. I have zero fear that anyone who likes my stuff is going to like an AI-generated piece of trash. And, on a free site, I don't see much harm. But it does get complicated, fast, where money or recognition are involved, and therein lies the problem; I get that. I'm glad I'm not in charge of figuring out the response here, but I am at my work. And it's a pain in the fuckin' ass, let me tell you.

 

[drobert44]

Honestly, I'm not sure it matters all that much.

AI produces shit stories. So do some human writers. This site (and many others) proves there's a market for shit stories, regardless of the navel-gazing of those of us in the AH (who genuinely do care about excellence in writing, I think). We're a tiny minority in a sea of preferences.

So it's not that clear to me how "dangerous" AI content is to those of us who plug away at this. I have zero fear that anyone who likes my stuff is going to like an AI-generated piece of trash. And, on a free site, I don't see much harm. But it does get complicated, fast, where money or recognition are involved, and therein lies the problem; I get that. I'm glad I'm not in charge of figuring out the response here, but I am at my work. And it's a pain in the fuckin' ass, let me tell you.

AI is scary. It aint human, but it sure moves fast. The next war will be fought with drones, and the best AI will win it..........

 

[EroticPower]

In your opinion. Based on your speculation.

Emily

You’re right, it is ALL based on speculation. But we’ve seen how AI has evolved over the years, from 2021, AI written stories were the meme of the internet with how bad they were. Now three years later, AI has gotten smarter and its result into this. Imagine what would happen in five-ten years. Imagine what would happen next. You say we should worry on today, not tomorrow but if that’s the case, then this AI rejection wouldn't be happening and this site would just be like the others where they just let AI stories in not worrying about the legal consequences in the future. Laurel is right to do this, as in the future, there might be some new laws regarding AI and it could affect websites like this one. But on the other hand, while there would be new laws established around this, it doesn’t stop AI from evolving, it’s just escalation and the problem will get worse for everyone until the government does something about it.

 

[EmilyMiller]

You’re right, it is ALL based on speculation. But we’ve seen how AI has evolved over the years, from 2021, AI written stories were the meme of the internet with how bad they were. Now three years later, AI has gotten smarter and its result into this. Imagine what would happen in five-ten years. Imagine what would happen next. You say we should worry on today, not tomorrow but if that’s the case, then this AI rejection wouldn't be happening and this site would just be like the others where they just let AI stories in not worrying about the legal consequences in the future. Laurel is right to do this, as in the future, there might be some new laws regarding AI and it could affect websites like this one. But on the other hand, while there would be new laws established around this, it doesn’t stop AI from evolving, it’s just escalation and the problem will get worse for everyone until the government does something about it.

So lobby your Congressman or woman.

Emily

 

[EroticPower]

Honestly, I'm not sure it matters all that much.

AI produces shit stories. So do some human writers. This site (and many others) proves there's a market for shit stories, regardless of the navel-gazing of those of us in the AH (who genuinely do care about excellence in writing, I think). We're a tiny minority in a sea of preferences.

So it's not that clear to me how "dangerous" AI content is to those of us who plug away at this. I have zero fear that anyone who likes my stuff is going to like an AI-generated piece of trash. And, on a free site, I don't see much harm. But it does get complicated, fast, where money or recognition are involved, and therein lies the problem; I get that. I'm glad I'm not in charge of figuring out the response here, but I am at my work. And it's a pain in the fuckin' ass, let me tell you.

Yeah I don’t envy Laurel going through all this,but I do wonder about other sites and why they don’t adopt the same policy. It just confuses me on that one as those sites could get in trouble if what Laurel is fearing comes true in terms of legal ramifications for AI.

 

[AwkwardlySet]

Ergo, expecting Literotica to "solve" AI is, in my view, naive.

I agree. But the AI problem has been around for quite a while now, and while it's quite unrealistic for Laurel and Manu to develop a solution, they should have developed a reasonable strategy by now, in my opinion.
Using some obscure AI tool and undisclosed criteria for AI rejections doesn't really help anyone. It certainly doesn't help those authors whose stories get wrongfully rejected with no pointers on what to do to get past the approval process. There are no official guidelines from Literotica for such authors, even if there are clearly so many of them. So they turn to AH where they get benevolent advice from us that might improve their writing, but it's impossible to know how much that advice really helps those authors pass the approval process, because we, same as them, have no clue what Laurel's criteria is.

For example, an imperfect but workable strategy would be to say: "We use the *insert name here* tool for screening submissions. We set our threshold at *insert percentage here* of human content for a story to be approved. Keep in mind that your story might additionally be screened by one of the website administrators even if it passes the tool test."

Such a strategy would, in my opinion, allow all the authors to test their stories BEFORE submitting them, thus making Laurel's job much easier, which would, in turn, let her dedicate some time to other aspects of running Lit that are heavily suffering right now. Instead, this chaotic approach is creating anger and frustration on so many levels and, as I claimed many times before, isn't getting punished only because there is no real competition for Literotica out there, and I say this by having tested other most popular free story sites.

 

[AG31]

If you don't want to read another post about AI rejection then move on. However, I think this adds to the conversation based on everything I have read on here. Warning, it is quite long when you include the four versions of my story intro (explained below).

I know many are dismissive about the importance of having conversations with people who have been rejected for using AI. Getting a story blocked from publishing on Lit might not be a huge deal, but as this article (https://www.fastcompany.com/91074029/can-using-grammarly-set-off-ai-detection-software_) shows, there are real consequences for getting it wrong.

I have my personal story (below), but at the very least, I think Literotica needs a clear and straightforward AI policy. If using Grammarly or others to edit spelling and grammar will trigger the AI detection, SAY THAT, don't pretend that it is okay, and then accuse writers of largely composing their stories using AI.

I cannot speak for anyone but myself, and I know many will not believe what I say here or find reasons to pick at my story, but I think it is important to share how I found myself getting rejected for "using" AI.

I started publishing on Lit in 2020 and have been reading stories here for almost as long as it has existed. My first stories were rough, with me struggling to self-edit (as many do). I quickly sought out volunteer editors, but after using four of them, long waits, ghosting, and unsatisfactory results discouraged me from continuing to use VE's. I turned to Grammarly, which, at the time, no one seemed to have any issue with on Lit.

Was it perfect? Of course not, but I was happier with the results of my work when I used it. I re-edited many of my stories using it and published an unfinished nine-part series under a different pen name using it and several stories under this one. I was happy, my readers seemed happy, and there were no problems getting anything published when I stopped publishing on Lit in late 2022. I recently returned to Lit and published two stories with no issues using the same editing process.

To be clear, my use of Grammarly has not changed over the years. I tend to write my first draft in Pages with spelling and grammar check turned off. I write faster this way and can get into a good flow. However, I make a lot of typos, misuse homophones, leave out entire words, and my comma usage becomes erratic at best. Needless to say, it is not publishable at this point.

I then turn on the crappy spelling and grammar check in Pages and do a read-through, correcting what I see and what is marked (that I agree with) by the program. This draft is usually better, but only slightly, because I truly do suck at proofreading my own work without it being pointed out.

My next step is to load it into the free web version of Grammarly and correct or ignore everything with a red underline. I will then give a final read-through, taking note of the yellow "premium suggestions" as areas to focus on. Now, for those that don't know, these are only marked on the free version, and there are no suggestions for fixing the issue, only that there might be an issue. What this does for me is let me focus on those spots for further examination and determine if I want to change them myself. For me, most of these are a result of passive voice or comma usage. I fix the ones I want and leave the others.

This is usually when I am happy enough to publish. At most, I will give another read-through to make sure.

Please note, that at no point in the process have I selected the "Generative AI" function or used any suggestions from Grammarly other than things like "peek not peak," "self-edit not self-edit," "inset comma," "remove comma," and run of the mill spelling that slipped through Pages inadequate spell check.

So, it is no surprise that I am confused when I see this message in my rejected story:

"Are you using Grammarly, ProWritingAid, Quillbot or similar software? Many modern writing packages incorporate AI. If you are using a grammar check program sparingly (as a spellcheck, to fix punctuation, review grammar, and/or occasionally as a thesaurus), that should be fine. If you are allowing a grammar check program to “rewrite” your words, that may cross the line into AI generated text/stories. Please see this FAQ for more information: https://literotica.com/faq/publishing/publishing-ai"

I read through the FAQ and came to the conclusion that I am not violating the AI terms. Especially since I have not changed my editing habits since before Grammarly introduced the new "AI" features. I promptly read through my story to see if anything stood out and resubmitted it with a few edits and a note:

"I am resubmitting this story because it was rejected for suspected use of AI. I have made some minor edits, as I decided to reread to see if anything stood out, but I can assure you that while I use the free version of Grammarly for spelling, punctuation, and other grammatical issues, I do not and have never used the generative AI function to create text for me. I have had the same editing process since at least my Late Night Truth or Dare series, including my two recent submissions. Thank you."

I should also point out that the story that was rejected, was the first part of the series I had previously published in 2022 under a different pen name and since removed.

I had two more stories in the queue, both of which were also rejected, as well as my resubmission of the first rejection, this time with a comment in response to my note of:

"Hello! We appreciate the effort you put into your work. In recent months, Grammarly has added generative AI to its product - so while previous works may have been not affected, any works you processed through Grammarly post-AI may introduce generative AI with its rephrasing features. We've checked this work several times and it is still coming back as being composed largely of AI-generated prose. Please see this FAQ for more information: https://literotica.com/faq/publishing/publishing-ai"

This is when I became angry. Not only were my assertions being dismissed, it was being blatantly stated that my work was "composed largely of AI-generated prose," by the AI that is reviewing for AI. I could understand if Literotica is taking a zero-tolerance approach and blocking anything that passed through Grammarly (which they are not saying), but to accuse me of using work generated by an AI, not just edited, makes me feel like I am being dismissed and tossed aside without being heard. I write my own stories.

It should be noted that a week prior to this second rejection of my story, I messaged @Laurel explaining my situation and providing the first 2500 characters of my new story in three different edits. I hadn't edited it before getting the rejection and reaching out allowing me to do this (I have deleted the first drafts of my other stories).

For those of you who doubt my story or think it is a matter of my "style." Please, read the four versions (I added one more than I sent to Laurel) of my intro below (unedited, edited in Pages, Only the red underlines in Grammarly, final Grammarly edit as described above). I know I am no great author, but I hope you can see that this can't be AI, if only because of the copious mistakes in the first draft and the lack of major changes in the last. Be honest, and think about whether this should be banned or not. I have not submitted it yet, but it is 100% representative of what I have done in writing and editing three banned stories and numerous published ones before those.

I have taken to time and risk to expose my writing process, showing the flaws of my early drafts, and if you choose to read them, I hope you take the time to think about the impact of the Literotica AI policies as they currently stand.

I have no influence on the workings of Lit, and I don't know if any of you do, but what I know is that when I read the threads about AI rejections I see many dismissive or downright mean comments towards people who are frustrated by being excluded from a system that has been so welcoming for so long. Is everyone telling the truth about their AI usage? I have no idea, but I am, and I am willing to bet most people are. There really is no reason to lie. In all honesty, there is no reason to try and publish AI-generated writing on Lit in my opinion (although I don't doubt it is being done). We are not making money and most of us simply want an audience and feedback. I am happy that people (mostly) like my stories, and I wouldn't get the same feeling if I had not written them myself.

I honestly don't know if I will respond to or even read the responses to this. I am tired, have other issues going on in my life, and just want to write as an escape. Literotica is no longer a happy place for me. I just wanted to say my piece and give the doubters of those struggling to figure out what's going on something more to think about. I think we deserve that.


For context, this is the start of Chapter 8 of a previously abandoned series of mine. The graduation being referred to is college, not high school, and of course, everyone is over eighteen, not that there is any sex in this part.

Thanks for this. It adds to the understanding of what's happening with AI checking on Lit. Dunno what the solution is.

 

[Voboy]

I agree. But the AI problem has been around for quite a while now, and while it's quite unrealistic for Laurel and Manu to develop a solution, they should have developed a reasonable strategy by now, in my opinion.
Using some obscure AI tool and undisclosed criteria for AI rejections doesn't really help anyone. It certainly doesn't help those authors whose stories get wrongfully rejected with no pointers on what to do to get past the approval process. There are no official guidelines from Literotica for such authors, even if there are clearly so many of them. So they turn to AH where they get benevolent advice from us that might improve their writing, but it's impossible to know how much that advice really helps those authors pass the approval process, because we, same as them, have no clue what Laurel's criteria is.

This is true... but it's also true to form. This is precisely how Laurel and Manu handle sweeps: an opaque process of great consequence, with endless AH threads full of well-meaning people giving unknowably credible advice. In the case of sweeps, the reason is that L&M don't want anybody finding out about the process enough to circumvent it somehow.

Is it such a leap to imagine they are handling this exactly the same way?

For example, an imperfect but workable strategy would be to say: "We use the *insert name here* tool for screening submissions. We set our threshold at *insert percentage here* of human content for a story to be approved. Keep in mind that your story might additionally be screened by one of the website administrators even if it passes the tool test."

Again, look at it from their perspective: this would increase the odds of some clever user defeating the process, whatever it is.

Such a strategy would, in my opinion, allow all the authors to test their stories BEFORE submitting them, thus making Laurel's job much easier, which would, in turn, let her dedicate some time to other aspects of running Lit that are heavily suffering right now.

Those of us who've been here for awhile are under no illusions that they're interested in fixing things that are "heavily suffering," because they do not seem to agree that that is the case. They are happy with the site as it is, and are demonstrably uninterested in making changes even if you (and others) deem them critically important.

The cost-benefit analysis you're making here is not, I think, similar to L&M's analysis.

I don't really disagree with you: I, too, would handle AI differently than L&M are doing it. But I think I'm empathetic enough to see it their way: they are worried about different things than you are, and that's nothing new.

 

[AwkwardlySet]

Again, look at it from their perspective: this would increase the odds of some clever user defeating the process, whatever it is.

I expected that argument. But what can really some potential AI user do to circumvent the screening - write a query to ChatGPT asking it to generate an AI story but in such a way as to avoid detection by that particular tool?
Laurel could tweak and test the human content percentage requirement until they are satisfied, but either way, I believe it would be more fair towards authors than what they are doing now. No criteria, no disclosure, no guidelines.

Those of us who've been here for awhile are under no illusions that they're interested in fixing things that are "heavily suffering," because they do not seem to agree that that is the case. They are happy with the site as it is, and are demonstrably uninterested in making changes even if you (and others) deem them critically important.

The cost-benefit analysis you're making here is not, I think, similar to L&M's analysis.

It took a year of my Lit experience at least before I realized how true that statement is, so I'll agree with you there easily, but I would like to believe that they want their website to run smoother, or at least as "smooth" as it ran before the AI problem blew up. Even I won't go that far in my criticism to claim that they hadn't been at least somewhat more responsive towards authors before the AI shitshow started.

 

[Bramblethorn]

AI is a problem. That problem is likely to get worse.

NO institution has figured out how to deal with that problem. Many of those institutions have vastly more power, money, time, and staff than Literotica does.

Ergo, expecting Literotica to "solve" AI is, in my view, naive. I think there will be many, many, many growing pains involved as we stupid-ass humans surrender so much control to these machines we're enabling to replace us.

I don't believe, with some optimists, that this "AI rejection debate" is a temporary problem with a solution likely to manifest itself soon. Rather, I think this is the new normal. It might be wise to start thinking about how to accommodate ourselves to that new normal.

I don't think we've reached the new normal yet, because the current business model for generative AI services is unsustainable. Generative AI is expensive to train and expensive to run. Although they're cagey with financial details, OpenAI seem to be running at a considerable loss for now, presumably in the hope of creating future demand and perhaps of killing off some of the competing options. (It's what Uber did: run at a loss for a few years, trying to drive old-school taxi services out of business, in the hope of exploiting a less competitive market afterwards.)

We're still in the "first one is free" stage of this story. Where it goes from here is anybody's guess, but it's not likely to stay the way it is. If it ends up with some of the current players going out of business and others switching from free models to paid/freemium, that might reduce the pressure on Literotica.

 

[scbolder]

The most typical rejection is something like: Lit: you used Grammarly suggestions quite a lot Author: didn’t realize they were AI-generated, will fix Lit: cool!

My interaction has been more Lit: you used AI. Me: No, but I did I used the spelling and grammar on Grammarly. Lit: We ran it multiple times and they all say it is largely written by AI. Grammarly uses AI, maybe you used it without knowing. Me: message Laurel explaining my editing process (as I did above). Laurel: has not responded for 2 weeks at this point.

The thing that irritated me the most is that when I got my first rejection notice I read over all the AI guidelines and knew that I had used Grammarly in a way that should be approved by their own statements (only for spelling and grammar as it say should be okay), and was for the two stories I had just published.

Now, did using Grammarly in this way cause me to trip the AI detection? Probably, based on what I've read in the forums, but it is frustrating to be told you are violating the stated rules when you know you are not. There is a difference between triggering the AI detection and violating the stated AI rules that Literotica has published. They should be clearer about the process and what they are actually rejecting you for (AI detection, not violations of the stated guidelines).

I hope they modify their guidelines and rejection notice to indicate that using Grammarly, or the like, is likely to trigger AI detection and such triggering is why a story is being rejected whether or not you violated the guidelines. It is a change that would have made me react differently. Instead of getting angry that I had not violated the rules and was still being punished. I would have immediately resigned myself to not be able to edit with Grammarly. I would have been annoyed about my already edited stories, but not angry because I was being told I was doing something I wasn't.

If using Grammarly is a no go, then I am fine moving forward without it. I am almost done editing my most recent story without Grammarly and incorporating some of the suggestions I have recieved. I will submit in a few to days. Hopefully it goes through, if not, I have no idea.

The lingering issue for me, at this point, is that I have 20 chapters in three different series that I wrote over the last two years but did not publish (or posted and removed from another pen name) that I would like to publish and get out there. The problem is that I edited them in Grammarly and did not save prior versions. At this point I have no idea if I can edit out the Grammarly influence enough to not trigger the AI detection. We will see, but I hope to not have to fully rewrite them.

 

[EmilyMiller]

My interaction has been more Lit: you used AI. Me: No, but I did I used the spelling and grammar on Grammarly. Lit: We ran it multiple times and they all say it is largely written by AI. Grammarly uses AI, maybe you used it without knowing. Me: message Laurel explaining my editing process (as I did above). Laurel: has not responded for 2 weeks at this point.

The thing that irritated me the most is that when I got my first rejection notice I read over all the AI guidelines and knew that I had used Grammarly in a way that should be approved by their own statements (only for spelling and grammar as it say should be okay), and was for the two stories I had just published.

Now, did using Grammarly in this way cause me to trip the AI detection? Probably, based on what I've read in the forums, but it is frustrating to be told you are violating the stated rules when you know you are not. There is a difference between triggering the AI detection and violating the stated AI rules that Literotica has published. They should be clearer about the process and what they are actually rejecting you for (AI detection, not violations of the stated guidelines).

I hope they modify their guidelines and rejection notice to indicate that using Grammarly, or the like, is likely to trigger AI detection and such triggering is why a story is being rejected whether or not you violated the guidelines. It is a change that would have made me react differently. Instead of getting angry that I had not violated the rules and was still being punished. I would have immediately resigned myself to not be able to edit with Grammarly. I would have been annoyed about my already edited stories, but not angry because I was being told I was doing something I wasn't.

If using Grammarly is a no go, then I am fine moving forward without it. I am almost done editing my most recent story without Grammarly and incorporating some of the suggestions I have recieved. I will submit in a few to days. Hopefully it goes through, if not, I have no idea.

The lingering issue for me, at this point, is that I have 20 chapters in three different series that I wrote over the last two years but did not publish (or posted and removed from another pen name) that I would like to publish and get out there. The problem is that I edited them in Grammarly and did not save prior versions. At this point I have no idea if I can edit out the Grammarly influence enough to not trigger the AI detection. We will see, but I hope to not have to fully rewrite them.

I’m describing the most frequent process, not edge cases, as I said. I know a number of people have got caught in the loop you describe. I’m led to believe it’s a minority.

Emily

 

[Voboy]

I don't think we've reached the new normal yet, because the current business model for generative AI services is unsustainable. Generative AI is expensive to train and expensive to run. Although they're cagey with financial details, OpenAI seem to be running at a considerable loss for now, presumably in the hope of creating future demand and perhaps of killing off some of the competing options. (It's what Uber did: run at a loss for a few years, trying to drive old-school taxi services out of business, in the hope of exploiting a less competitive market afterwards.)

We're still in the "first one is free" stage of this story. Where it goes from here is anybody's guess, but it's not likely to stay the way it is. If it ends up with some of the current players going out of business and others switching from free models to paid/freemium, that might reduce the pressure on Literotica.

I don't think the "new normal" is free AI.

I think the "new normal" is iterative AI that is not easily detectable.

 

[BobbyBrandt]

If you don't want to read another post about AI rejection then move on. However, I think this adds to the conversation based on everything I have read on here. Warning, it is quite long when you include the four versions of my story intro (explained below).

I know many are dismissive about the importance of having conversations with people who have been rejected for using AI. Getting a story blocked from publishing on Lit might not be a huge deal, but as this article (https://www.fastcompany.com/91074029/can-using-grammarly-set-off-ai-detection-software_) shows, there are real consequences for getting it wrong.

I have my personal story (below), but at the very least, I think Literotica needs a clear and straightforward AI policy. If using Grammarly or others to edit spelling and grammar will trigger the AI detection, SAY THAT, don't pretend that it is okay, and then accuse writers of largely composing their stories using AI.

I cannot speak for anyone but myself, and I know many will not believe what I say here or find reasons to pick at my story, but I think it is important to share how I found myself getting rejected for "using" AI.

I started publishing on Lit in 2020 and have been reading stories here for almost as long as it has existed. My first stories were rough, with me struggling to self-edit (as many do). I quickly sought out volunteer editors, but after using four of them, long waits, ghosting, and unsatisfactory results discouraged me from continuing to use VE's. I turned to Grammarly, which, at the time, no one seemed to have any issue with on Lit.

Was it perfect? Of course not, but I was happier with the results of my work when I used it. I re-edited many of my stories using it and published an unfinished nine-part series under a different pen name using it and several stories under this one. I was happy, my readers seemed happy, and there were no problems getting anything published when I stopped publishing on Lit in late 2022. I recently returned to Lit and published two stories with no issues using the same editing process.

To be clear, my use of Grammarly has not changed over the years. I tend to write my first draft in Pages with spelling and grammar check turned off. I write faster this way and can get into a good flow. However, I make a lot of typos, misuse homophones, leave out entire words, and my comma usage becomes erratic at best. Needless to say, it is not publishable at this point.

I then turn on the crappy spelling and grammar check in Pages and do a read-through, correcting what I see and what is marked (that I agree with) by the program. This draft is usually better, but only slightly, because I truly do suck at proofreading my own work without it being pointed out.

My next step is to load it into the free web version of Grammarly and correct or ignore everything with a red underline. I will then give a final read-through, taking note of the yellow "premium suggestions" as areas to focus on. Now, for those that don't know, these are only marked on the free version, and there are no suggestions for fixing the issue, only that there might be an issue. What this does for me is let me focus on those spots for further examination and determine if I want to change them myself. For me, most of these are a result of passive voice or comma usage. I fix the ones I want and leave the others.

This is usually when I am happy enough to publish. At most, I will give another read-through to make sure.

Please note, that at no point in the process have I selected the "Generative AI" function or used any suggestions from Grammarly other than things like "peek not peak," "self-edit not self-edit," "inset comma," "remove comma," and run of the mill spelling that slipped through Pages inadequate spell check.

So, it is no surprise that I am confused when I see this message in my rejected story:

"Are you using Grammarly, ProWritingAid, Quillbot or similar software? Many modern writing packages incorporate AI. If you are using a grammar check program sparingly (as a spellcheck, to fix punctuation, review grammar, and/or occasionally as a thesaurus), that should be fine. If you are allowing a grammar check program to “rewrite” your words, that may cross the line into AI generated text/stories. Please see this FAQ for more information: https://literotica.com/faq/publishing/publishing-ai"

I read through the FAQ and came to the conclusion that I am not violating the AI terms. Especially since I have not changed my editing habits since before Grammarly introduced the new "AI" features. I promptly read through my story to see if anything stood out and resubmitted it with a few edits and a note:

"I am resubmitting this story because it was rejected for suspected use of AI. I have made some minor edits, as I decided to reread to see if anything stood out, but I can assure you that while I use the free version of Grammarly for spelling, punctuation, and other grammatical issues, I do not and have never used the generative AI function to create text for me. I have had the same editing process since at least my Late Night Truth or Dare series, including my two recent submissions. Thank you."

I should also point out that the story that was rejected, was the first part of the series I had previously published in 2022 under a different pen name and since removed.

I had two more stories in the queue, both of which were also rejected, as well as my resubmission of the first rejection, this time with a comment in response to my note of:

"Hello! We appreciate the effort you put into your work. In recent months, Grammarly has added generative AI to its product - so while previous works may have been not affected, any works you processed through Grammarly post-AI may introduce generative AI with its rephrasing features. We've checked this work several times and it is still coming back as being composed largely of AI-generated prose. Please see this FAQ for more information: https://literotica.com/faq/publishing/publishing-ai"

This is when I became angry. Not only were my assertions being dismissed, it was being blatantly stated that my work was "composed largely of AI-generated prose," by the AI that is reviewing for AI. I could understand if Literotica is taking a zero-tolerance approach and blocking anything that passed through Grammarly (which they are not saying), but to accuse me of using work generated by an AI, not just edited, makes me feel like I am being dismissed and tossed aside without being heard. I write my own stories.

It should be noted that a week prior to this second rejection of my story, I messaged @Laurel explaining my situation and providing the first 2500 characters of my new story in three different edits. I hadn't edited it before getting the rejection and reaching out allowing me to do this (I have deleted the first drafts of my other stories).

For those of you who doubt my story or think it is a matter of my "style." Please, read the four versions (I added one more than I sent to Laurel) of my intro below (unedited, edited in Pages, Only the red underlines in Grammarly, final Grammarly edit as described above). I know I am no great author, but I hope you can see that this can't be AI, if only because of the copious mistakes in the first draft and the lack of major changes in the last. Be honest, and think about whether this should be banned or not. I have not submitted it yet, but it is 100% representative of what I have done in writing and editing three banned stories and numerous published ones before those.

I have taken to time and risk to expose my writing process, showing the flaws of my early drafts, and if you choose to read them, I hope you take the time to think about the impact of the Literotica AI policies as they currently stand.

I have no influence on the workings of Lit, and I don't know if any of you do, but what I know is that when I read the threads about AI rejections I see many dismissive or downright mean comments towards people who are frustrated by being excluded from a system that has been so welcoming for so long. Is everyone telling the truth about their AI usage? I have no idea, but I am, and I am willing to bet most people are. There really is no reason to lie. In all honesty, there is no reason to try and publish AI-generated writing on Lit in my opinion (although I don't doubt it is being done). We are not making money and most of us simply want an audience and feedback. I am happy that people (mostly) like my stories, and I wouldn't get the same feeling if I had not written them myself.

I honestly don't know if I will respond to or even read the responses to this. I am tired, have other issues going on in my life, and just want to write as an escape. Literotica is no longer a happy place for me. I just wanted to say my piece and give the doubters of those struggling to figure out what's going on something more to think about. I think we deserve that.


For context, this is the start of Chapter 8 of a previously abandoned series of mine. The graduation being referred to is college, not high school, and of course, everyone is over eighteen, not that there is any sex in this part.

My process isn't much different than what you describe when writing for Lit, except that I use MS Word and the Grammarly plug-in feature instead of their website. I have never had a rejection.

Out of curiosity, how do you submit your stories once they are edited to your satisfaction? I always upload a Word document instead of using the submission text box and I wonder if that influences the AI detection process when Grammarly or similar tools are used. Maybe others who have faced AI rejections could share their submission process as well. Could the metadata embedded in an uploaded file prove authenticity

 

[MillieDynamite]

I've run AI text through multiple Plagiarism checkers and never had a flag go up other than a random line or two. The AI may scrap things from any source out there but doesn't copy and paste things. The damn things rework what they steal and make it some lesser than it was. Yes, it may steal words, but doesn't mean it's stealing words. If it is stealing isn't doing a good job of it. It bloats the sentences and paragraphs until they aren't flowing well. They may not be passive but they aren't moving like good literature either. There is no timing in AI work, not building up to this or that. Not any that I can see.

I have looked a lot into this, scoured a lot of the forums complaining about AI, and it seems to revolve around the fear of being in legal trouble because AI is basically using stolen content from other stories and just preparing just in case there is any blowback from it. But I feel that AI assisted and AI Generated stories have their own section and I think the author of that story should make sure to let their readers know that AI was involved in the writing, that way people can make an informed decision whether they want to read it. I know someone has made that idea in one of the forums and it received a mixed reaction but I do feel that would be the next logical solution going forward.

Because right now, it’s a s**t show. People’s work are getting rejected for AI even though they claim it’s not AI (maybe some of them are lying but the thing is the owner are putting all of them in the same basket using a shoddy AI Detection software to detect AI Written work that’s been established that AI detection software do not work, they are the lie detectors of the 21st century)). And its now gone to the point where people are being asked to change their writing style as it looks similar to how AI does it which justifiably they would be angry and frustrated that they are being asked to change their style, their way of writing,(even though as we established AI didn’t create that writing, it copies other people’s work, meaning it uses another person’s writing style and what’s worse, even though we can spot the telltale signs that it’s written by AI, one day it won’t be that easy anymore, it will come to a point where it would be indistinguishable). This policy is not a long term solution as all it does is push people away, alienate potential authors who could have contributed to this community.I understand there are people who in this community who see this as a good thing judging from reading the forums, and that’s fine as I agree with you to a certain point. Anyone who uses AI to write their own stories (not assisted writing) are not authors of their works, they are basically making Frankenstein and the stitches are more easier to see at the moment, one day it won’t. And the community that was once a friendly place is now a shadow of its former self, people attacking each other, accusing each other, belittling, bullying, these problems existed before the AI thing but now it’s exacerbated this problem.

I worry for this site’s future, not because it’s all relying on someone who has probably worked non stop, for the past 2 decades keeping this site up,but also the fear that if there is blowback if will hit the owner with legal ramifications that will shut the website down as a result (Which I understand that fear as this is her creation) but that fear has resulted in going in the complete opposite where now people who don’t use AI are being rejected and put in the same basket as the people who use AI so now they feel like they do not want to write anymore or like I’ve seen on the forums people recommend to try anywhere else(which while it’s a short term solution for now, like I said, when AI evolves to the point when their writing is indistinguishable, more authors are going to have their work rejected, pushing more and more people out).

In my opinion, the sites that does need to worry about legal ramifications is sites like Amazon and DeviantArt as they let AI work be monetised there, that is wrong as the thief is profiting off stolen bits of work, that’s where I draw a line, that’s when it goes too far when it is monetised.

For authors who have their work rejected, maybe it’s best to wait till the AI fiasco dies down within a year or two, in the meantime, I recommend trying out alternative sites instead like Archive Of Our Own or maybe write your own stories in your own blog or website, make your own corner where you can express your own stories maybe, that’s up to you. But I believe within a year or two, there will be a better solution that will suit the writers,the moderators and the owner of this site because at the end of the day, we want the website to keep continuing as it’s stacked with hundreds of stories, stories that people love and remember.

-For now,

 

[Paul_Chance]

As a web admin or system admin you do not publicly disclose the tools used to screen or filter content. That's just bad business. The bad actors then use that information to try and bypass the tools. The same would hold true for any counter AI tools or filters in place.

 

[StillStunned]

Maybe we've got it all wrong. Maybe, after 25 years, Laurel is fed up of reading bland erotica. Now she wants to read the weird stuff. To see what we can come up with as to prove we're human as the AI hammer comes down harder and harder.

 

[Paul_Chance]

Maybe we've got it all wrong. Maybe, after 25 years, Laurel is fed up of reading bland erotica. Now she wants to read the weird stuff. To see what we can come up with as to prove we're human as the AI hammer comes down harder and harder.

Tentacles for everyone!

 

[StillStunned]

Tentacles for everyone!

I'll write about Eleventacles. That's right, they go all the way to eleven!

 

[scbolder]

Out of curiosity, how do you submit your stories once they are edited to your satisfaction?

From Pages I save it as a plain text document and then upload that during the submission process. I don't think this has anything to do with it, but you never know.

 

[BobbyBrandt]

From Pages I save it as a plain text document and then upload that during the submission process. I don't think this has anything to do with it, but you never know.

Plain text would typically strip any metadata. I would still be curious about whether anyone who had a story where Grammarly was used for editing saw their story rejected when submitted as a *doc file.

 

[amenarimix]

I'm sorry you are having this problem with the Site. I believe your story. But having read the portions of your story in various versions, I think I understand why the Site is doing what it's doing. To me, your story reads like it's AI-generated. Your sentences are repetitively over-long, for one, with too many phrases and clauses strung together. Mix long sentences with short ones. You put phrases and clauses together that don't quite make sense to me, such as in the very first paragraph, as StillStunned pointed out. There's too much explanation. It doesn't feel quite right.

I agree with all of StillStunned's comments on your text.

The dialogue reads funny to me. The sentences are too long, and people do not speak in the real world in long sentences with perfect grammar. Insert errors into the dialogue. Make sentences shorter and more staccato. Add human filler words like "like" and "you know" into the dialogue.


This makes me wonder if what's happening is that AI is influencing what people think is "correct" prose, and people are starting to emulate it. Those of us who wrote before AI don't write in AI style and may not be having the problem to the same degree.

You know AI use 'you know and Like' very often? But you're right about that too many clauses. I used to take chatgpt help to express characters emotions. Always writing he said, she said and replied is boring. There I noticed AI like to use as many clauses as possible in a sentence.

To avoid AI detection we all need to put intentional errors and short simple sentences. Back in the day my readers used to eat me alive because of errors.

I really understand this writer story because I was just like him. Volunteer Editors are the worst except a few of them.

Anyway it would be best if literotica told us what AI tool they use. We could check ourselves and fix the issues.