I think he was referring to the old rules before it changed, as much as it felt frustrating at the time (sometimes), the one thing it did was stop The issues that was their before March , when there were a lot of AI content that I would admit was very very bad to the extent i remember people was even leaving the AI memory summaries of what the AI produced in the text in the stories . There is nowhere near as much AI stuff as people think , I had a chapter posted today but I counted 12 stories before the next one with an AI tag 
2 Likes
Hi hello, I don’t have anything to add to this discussion because I will probably be blocking the new tag anyway, but when I tried to look at the hyperlink where the new rules about this topic are supposed to be (the one in the announcement and in the first post on this topic), it says the page doesn’t exist/is private. Is this something on my end, or is the page gone?
Thanks for the notice. It seems I accidentally clicked the wrong button.
The posting is restored again, and the link should work now.
P.S.: It’s your prerogative to block the tag, of course, but I’d like to state that you’d be the only one who’d lose because of that. Generalisation is never good, and not all stories which were created with any kind AI involvement are automatically bad and not worth being read.
1 Like
Thank you for making the page viewable again. I took a look at the new rules and they seem clear, even if I personally disagree with them. You say that I shouldn’t generalise, but doesn’t the use of 1 tag as opposed to multiple generalise the use of AI in these stories?
As for your comment about me probably blocking the tag, I appreciate you standing behind your own rules and views, but I’ve got my own as well, so I’ll be blocking the tag regardless. Without going into too much detail, I work in the psychological field, and both Gen AI and AI chat bots like ChatGPT have negatively affected my work and the people I work with, so my view of AI of any kind obviously isn’t good.
Besides that, the biggest problem I have with AI, even if it’s only using it to help with spelling or other small things, I think any amount of use takes away something from being truly, fully ours. I’d personally rather read something that’s raw and 100% their own, than something that they needed help with from a machine that, as of now, has no soul to speak of itself.
This is just my own opinion, of course. Anyone is free to do what they want. These are my own choices/morals and I stand by them, so I won’t be losing anything in my eyes.
No, as the stories which would previously need to be tagged as “AI created” are no longer allowed on the site at all. So I don’t see the need for multiple tags any longer. What would they discern?
Then you must never read any book. Because all books are going through a long process of editing and reediting by usually more than one person, and these days also digital assistants, so it’s never entirely just the author’s voice. And never was. Even the old Greek texts were changed, translated, and altered throughout time.
As a non-native speaker, I’m glad that I can resort to tools like Grammarly to fix my grammar and spelling mistakes. As for the phrase optimizations, it suggests… sometimes I take them, sometimes not, as I agree that following them all the time CAN destroy your own style.
But understand that not every author is this competent and has the experience to develop a recognizable voice and style. They still might have a vivid imagination and an enticing story to tell.
Whether someone SHOULD better try to find their own voice over time or resort to mechanical help is another issue altogether. But we’re not a literature and writing school, we’re a smut site.
2 Likes
Thank you for clarifying the tag rule, I must’ve misunderstood that part.
And thank you for sharing your viewpoint on what I said. I can definitely understand the points you make, but the way things are right now, I won’t change my viewpoint on it. I hope you can understand and/or accept that. So far, I’ve been fine blocking the old tags, so I’m sure I will be fine blocking the new one as well.
1 Like
The tag exists precisely for that reason, it’s the only reason for its existence, a way for authors to choose to give readers who dislike texts where some artifacts of AI generation still remain, a possibility to opt out, or maybe take a risk. That is why we will detect those artifacts and encourage the author to tag their story when they are present.
Bear in mind that the tag only applies to texts where those artifacts (little bad habits of Gen AI) are detectable. But then, if they aren’t, the text won’t have much in its content to distinguish it from human writing. As someone who detects those artifacts, I can tell you that when you reach a point where they are no longer present, it means the author did much more than light editing, and transformed and reshaped the text so much that the initial AI input is barely there anymore, and that probably takes more effort than actually writing the text from scratch in the first place.
My point is, it would be ridiculous to add this tag, spend a lot of effort to detect when the author should be encouraged to use it, and then be mad that you do use it. So by all means, you do you.
As someone who writes a much less popular type of stories than what’s usually published and sought here, I can relate to your fears. Each of my stories compete for attention with 15-20 stories about more popular kinks.
But at the same time, do they really? Readers know what they want. If my stories are what they like, they’re pretty good at sifting through what they don’t want. With a search engine and ratings and comments and tags, it’s rather easy to make choices here.
To use your own analogy, I can see a lot of gallery goers just going past all the slop and hone in towards the good stuff. People will recommend it or they’ll just sigh at the men with six fingers and the weird ears, and hunt for something better.
A story is either better than other stories or it’s not no matter how stories are created.
If it is, then there’s no competition. People will flock to it because there are many ways to find it on site like here. And if you write great stories, why would you limit where you publish? The more places it appears in, the more potential readers you reach. If you say, I don’t want to publish on Site A because its policies limit its potential readership, I will instead publish on Site B, why not publish on Site A AND Site B? The number of readers you’ll get on Site A won’t be negative, it can only add to the total.
If it is not better than other stories, or to be more precise, struggles to find an audience who find it good, then the problem is not really the presence of competition, is it? Would the solution really be to prevent better stories from being published?
In any case, I don’t think we’re at the stage yet where AI can actually write better than humans by themselves. The only way to get a good story with AI is if a human uses it only in a minimal capacity for suggestions, research, brainstorming, to see options from which to choose, but actually makes the most important choices. And that’s not likely to change soon despite the hype with AI. You’d be surprised at how limited AI actually is. Keep in mind that Sam Altman announced recently that they achieved a major victory in ChatGPT and that now it will no longer use em-dashes when the user requests it not to. Wow. Next stop, Shakespeare.
At this stage, we’re far from flooded with AI stories on GSS. It would be difficult to flood us now that we won’t accept anything that was produced with low effort.
So, you can decide what you want for where to publish your stories, and I understand the anxiety, but we’re not at this stage yet for sure.
3 Likes
After reading through all the discussion, it seems like some people are talking past each other without obviously recognizing the points. There is a huge difference between generative AI and the AI that people have used for decades when even typewriters had spell-check. That spell-check is AI. That is enormously different from typing into a prompt, “Write me a sexy gay erotica with nipple torture and dildo fun.”
I think that the former – the basic things we use AI for every day, even in this reply where I had a typo and my Mac auto-corrected it – is always going to be fine on this site and shouldn’t need to be tagged. I defy anyone to say they don’t use any splell-check.
In my opinion, for people for whom English is not their first language, there should be some other tag, and I really did like the old “ai-assisted” one. That would similarly apply to folks who don’t necessarily write so well in English normally and use tools like Grammarly.
Things like M$ Copilot (which I don’t use at all, am on a Mac, and have all the M$ tools and tracking like that disabled) that suggest different phrasing, to me, fall under the above paragraph of “ai-assisted:” You wrote the original idea, you have used a modern, machine-learning-based large language model to tweak that. Versus a basic dictionary lookup run by a computer to auto-correct “birdd” to “bird.”
In my opinion, the implication of the new “ai-prose” – or certainly what I infer from the term without reading the rules (which I have read) – is that it was written by AI without much human input. And I think what pretty much everyone is saying is that that’s not what we want: Some don’t necessarily want to read that nor compete with it, and folks who use assist tools to help their writing don’t want to be lumped into that implication that their ideas didn’t come from them because “ai-prose” could imply that.
Meanwhile, a story I’m actively writing but not submitting here (because while it is gay and erotica, its story is much more than the erotica part and folks come here more to get off than to read stories with plots that every few chapters have a sex scene), I have used ChatGPT to bounce ideas around for the underlying physics and ontological debates. I used it to tell me the maximum theoretical kWh battery that could fit into a bracelet and whether that had enough power to run the minimum theoretical holographic projector. I used ChatGPT to give me an astrophysical phenomenon that could produce significant enough antimatter to harvest for matter-antimatter reactions that was also safe to harvest, including throwing my own ideas in there to see if it was theoretically possible. None of that prose it spit out will be part of my story. The underlying concepts that I asked about and it suggested will be. I, personally, don’t think I need to label that story as even “ai-assisted” because it didn’t participate in the writing nor even the concepts at all, it just helped check the math because I didn’t want to spend days pouring through my college textbooks.
On the other hand, the new way the rules read to be enforced is that this is a losing battle and acknowledgement of that. Authors need to self-tag because the mod team can’t often tell anymore: They might think it is when it isn’t and the author could get super-pissed as a result, or the author might flat-out deny it when they did. The approving team doesn’t want nor need that headache, it seems, and I can certainly empathize with that.
So, it doesn’t seem like there’s any perfect solution. But, it does seem like a take-home from this thread so far is that @Corin might re-consider the “ai-prose” tag to something more indicative of assisted-by rather than written-by. Another potential option is for there to be another box when submitting a story that we could check options for:
- No AI assistance.
- AI help with ideas but no prose generation.
- Grammar and/or translation AI assistance.
- AI help with some phrasing.
- [note in this area] Any significant ideas/writing done by generative AI are not permitted on this site.
Whatever box(es) was/were checked would then be displayed with the story and filterable by readers. Perhaps those changes could address most of the concerns while also being a minimum additional effort on everyone’s part?
1 Like
The tag “AI Prose” will only be used when a significant amount of the text is clearly Gen AI. That’s why we switched it from “AI Assisted” which is broader as it could cover a lot of other uses of AI that we don’t want tagged. Authors are still free to put a disclaimer in their stories to disclose more subtle uses of AI but the tag AI Prose is only for stories in which it is obvious so that readers who dislike this type of writing can filter it. That’s it. As we mentioned, we are not focusing on how stories were created, but on what they contain. If you read the rules, the third type of story is not tagged or even encouraged to be tagged. But authors can still put a disclaimer if they want. We don’t want to be the police of AI.
2 Likes
Only asking this question to satisfy my curiosity as I wouldn’t stop using the AI tag regardless out of respect for the community even if the rules were different. But say you write a complete story 5000 words. You put it into a Ai tool a paragraph at a time, nothing about the story is changed , the dialogue is pretty much the same , all that Happens is the paragraphs in the story is rephrased or expanded . You end up with 6000/ 7000 words, which is still then edited there after . I assume it should be tagged AI because the rephrasing may still turn it into a AI type story?
I would be interested in the response to that.Because that is the sort of thing that I don’t see as AI wrote the story. If everything is getting expanded rather than just rephrased, then I might wonder about how much was a I.Assisted and perhaps written. But I think that it would be appropriate to do that. I don’t know that I agree that someone using it because it’s not their first language and they’re using it to translate that that then should be called written by an a I i do read stories that are written in other languages.Whether or not they were translated when put on the site or not.But I appreciate knowing going in what language they started with.Because many times that’s going to do something to that translation. Some translations are good, but many times there are words that don’t make sense or homonyms or something that got used.And shouldn’t have been, and that’s not the fault of the author really. But it is something that would be helpful to know.
I did like the comment that someone made about a I pros being versus a I assisted. And for me, that does hit a lot of the nail on the head.I don’t want my work being determined to just be a I when it wasn’t. And I probably have picked up some words or phrases from a I that I probably insert into my own writing now. But I can guarantee there are some phrases that chatgpt uses constantly that tick me off every time it happens.And I have to rewrite them because I don’t want them and I assume that may be some of what the detector finds.
Anyway, I appreciate this conversation and that more people are getting involved in it.Because I think authors and readers need to be part of this conversation.
1 Like
As mentioned earlier, the process you use to make the story is not important. What’s important is the result. So depending on how much you edited the text, it could still be full of AI artifacts which are easily recognizable in the prose, or they could have been completely removed. In the first case, we would encourage the tag. In the other case, we wouldn’t be able to notice, so we wouldn’t ask for a tag although you could decide to disclose your specific use of AI in a disclaimer at the top of your story.
2 Likes
Also, we can’t put a life story into two or three words for a tag.
”AI Prose” does not mean that the story is AI Prose, but that the story contains AI Prose. In the same way all tags work. A tag of “Oral” does not mean that the story is all one big blowjob, but that the story contains oral sex.
What is meant is: “Contains some AI-like Prose” or “Contains some text similar to what GenAI produces” but those are mouthfuls. We haven’t yet found the best way to express it and this is so complex a subject that we might never find a non verbose way of describing it.
That’s the only aspect of AI use that can be detected somewhat reliably by either readers, approvers or experts, although still never provable beyond a doubt. And the only aspect that is observable in the content of the story and not about the black box of the creation process behind it. So the only thing that makes sense to tag.
The main reason for the change is that something like “AI Assisted” puts the focus on how the story was created instead of what we see in the story, what it contains, and that also there are a million ways in which a story might have been assisted by AI, so we didn’t want people to tag just because they asked a question to ChatGPT in making the story.
1 Like
Im happy with those answers . Thank you 
to be honest I don’t really have a clue about artefacts or the type of writing AI produces outside of what I learned since I started submitting stories , which Is sometimes AI will use certain words a lot . Ie stark contrast gate lol . But then people say that phrase in audio books I’ve listened to so I guess part of the AI thing is repetition. The thing Im struggling with at the moment is conveying how words are spoken , feels like people are either murmuring , mumbling, whispering or purring lol 
which again Is maybe another AI giveaway , anyway I know this is all besides the point. Thanks for the clarity . The most important thing to me is not being in danger of my stories being on the banned list
Case #2 “Required” vs. “Encouraged” is a pretty big negative change IMO.
1 Like
That’s why I’m wondering if perhaps adding a short set of checkbox options when the author submits the story/chapter could be beneficial. It can capture those subtleties more than “ai-[one word]” per a tag, and it could be added to the filtering that readers can use on the site. Yes, it would need to be self-reported for the most part, but I think y’all should consider my suggestion there a bit more. And, from reading a few of the commenters on this thread, I think they would really prefer to be able to select that they used AI in a translation process and not get lumped into the implication / inference by other people with “ai-prose” that it was not all their original ideas.
Maybe I’m reading too much into others’ comments there, but yeah … something like a list of checkboxes when we submit the story to indicate no AI or how AI was used could capture those subtleties and could be used as filters by others.
Again, we don’t focus on how the story was created, but what it contains. For translations, if the author manages to prompt correctly and not get a text with AI artifacts, no tag will be needed. If it does contain text that reads like AI text, i.e. AI prose, then the AI Prose tag will be encouraged so that people can filter it. Doesn’t matter if you got the AI-like text by translation or by being a human who just happens to write like an AI. Readers will see that sort of text and not like it. It’s a convenience to readers we encourage authors to do. If they don’t, then a few readers will lose a minute or two figuring out they don’t like that writing and that’s all. They’ll read another free story among the tons of free stories on this site. Feeling entitled to a complete AI profile of every story in the site, apart from being impossible, is more than looking a gift horse in the mouth, it’s insisting in giving it a colonoscopy.
Authors can write a full paragraph in their disclaimers giving the full history of AI use from when they dreamed of using AI one day to what their last prompt was, if they feel so inclined.
The rules as they are now require a lot already from our unpaid resources.
1 Like