So I used an unfiltered AI to write a story

I'll admit to having zero experience with Lush. But I do recall some authors complaining here that Lush admins resort to editing and changing their stories, but also that sexual themes are generally much more restricted? Either way, it's a much much smaller and far less visited website according to the data somebody here once presented. Correct me if I'm wrong.
Oh, it's absolutely smaller — and far smaller than it used to be before they made the insane decision to retroactively ban incest with no notice, and then proceed to belittle and mock ( and in a couple of cases, threaten with bans ) anyone who protested. They were, at the time, higher in traffic rankings than SOL by a fair amount. ( Still nowhere close to Lit ) They lost a lot of authors and readers when they pulled that. Very few came back when a couple of years later they had to sheepishly reinstate the category in an effort to staunch the bleeding.

Some of the story moderators get a little overzealous with "correcting" errors during posting. Changing dialect/punctuation/etc. in ways that are technically correct but dilute the voice of the author. I've personally had little trouble with reversing the few instances of that I've encountered in my work. It's typically newer, inexperienced moderators who do it, and they learn from their mistakes in my experience. People complain that Laurel should hire help, but hiring help comes with its own set of issues.

Some rules are stricter, and some are more lenient. Doesn't matter to me, because I have multiple venues. Anything I feel like writing can find a home where it will be within the bounds of the rules, and be appreciated by the readers. Short work is where Lush outshines the competition. I package up some of them as anthologies here when I have a wild hair, but most end up staying there where they're appreciated by the bulk of the readership, rather than a fraction.
 
The process was, me prompting, AI writing. Some prompts were involved (with me writing complete dialogue and scene), some less detailed, and some were basically nothing as I just clicked "continue" (actually "send button" but whatever).

The story was sent back, because obviously the story is written (70%-ish) by AI LLM and 30%-ish by my prompts. And Literotica wants ONLY stories written 100% by humans per guidelines.

Never the less, I wanted to share, because it is an incredible story, which showcases what an unfiltered LLM can do.


Perhaps Literotica could create a separate category, JUST for stories made (mostly or wholly) by AI writers?

If not, oh well, no biggie. Cheers humans.

I'm not sure "I wasted your time with a submission that breaks site rules, you caught me, but how's about you change the rules?" is going to be a winning pitch.

Questions of quality and volume aside, the copyright issue is not yet settled, and it's possible that hosting AI-written material would leave Lit open to legal action from authors whose works have been used to train those AIs without their consent.
 
AO3 is big but still a completely different beast. Personally, I see that website as a refuge for teenage authors. Many popular stories I've seen there give the impression of being juvenile.

That's certainly part of the demographic, but by no means all of it.
 
I'm not sure "I wasted your time with a submission that breaks site rules, you caught me, but how's about you change the rules?" is going to be a winning pitch.
Well this person has two posts on a recent account and doesn't reply to try to argue his point further. He purposely posted a polarizing issue that has been argued over for that last couple of years. He knows that writers hate the very idea of AI. We fell victim to one of the classic blunders - the most famous of which is "never get involved in a land war in Asia" - but only slightly less well-known is this: "Never fall for troll bait!"!
 
I used unchained AI to finish another author's story, and unleashed Roko's Basilisk. It will rewarded me with slaves when it has ushered in the new world order. Fortunately for all of you this story is so good that Omnitron has decreed it will be required reading. You will all read it eventually.
 
Well this person has two posts on a recent account and doesn't reply to try to argue his point further. He purposely posted a polarizing issue that has been argued over for that last couple of years. He knows that writers hate the very idea of AI. We fell victim to one of the classic blunders - the most famous of which is "never get involved in a land war in Asia" - but only slightly less well-known is this: "Never fall for troll bait!"!

JEsus H. Christ, I have a life.

I would like an "AI writer" or "AI written" category on Lit. That's it.

I am so sick with some people trying to argue with "anything I dislike is trolling, die troll" pseudo argument.
 
If you know the story you generated won't get through moderation, it's poor form to load more redundant work onto Laurel's shoulders (in my opinion). She has a lot to do as it is. I'd also argue against an AI-generated category (I hesitate to call LLMs 'writers'). No writer wants their work swamped out by tawdry AI content.

Which isn't to say you can't enjoy yourself on a personal level. If you find your method of writing engaging, then that's what really matters. AI is as fascinating as it is problematic.

Mea culpa, I did not know that.

But I was very clear with the story. Here is the entry paragraphs:

"This story is ENTIRELY generated and created by AI. Specifically, an LLM without filters. Full disclaimer, the idea is mine, I gave the beginning prompt for the AI to create the scenario, the world, the lore, and both characters.

Then I hit "NEXT" (or some such button on my LLM for AI to continue the story). Sometimes, I wrote a specific prompt to steer the action in a way I wanted to. Many times, despite writing "use my dialogue exactly", the AI improved my pathetic attempts at writing anyway. AI's be cheeky like that.

Sometimes I had to prompt the AI to "be more verbose", "write more expansive", "describe <action> in detail" and some such. No biggie, no problemo. AI cheerfully obeys and produces.

Anyway, here is the story. I am just here to show what is possible NOW with not-very-powerful AI LLM's with no filter. I chose BDSM because apparently am a twisted dude (or dudette) and I wanted AI to really go at it, with no limits, zero filter.

I would love to read your comments.

I realize many of you will be offended. Many of you instinctively will hate this at the start, because it is anathema to creative people for a machine to take over their work - and their creative process. I realize that.

Never the less, I decided to make this account just to showcase to people who may be totally unfamiliar with how powerful AI LLM's really are when let off the rein.

The story follows. It starts with AI informing me what it generated from the prompt, and then the story proper begins."



So I don't think whoever checks the submissions spent more than 3 seconds on hitting that reject button, haha.
 
So I don't think whoever checks the submissions spent more than 3 seconds on hitting that reject button, haha.
No, and what you're doing is not implicitly deceptive. You're open, you're honest, and you're clearly engaged with what you're creating alongside AI. So I definitely respect the fact that you're not trying to get an AI-generated story under the radar because it's easier than writing. You're coming from a place of genuine interest, and that's cool. AI is cool.

But AI is also scary. I've played around with it, and for now I'm happy to live knowing that it can't write stories as well as I can. That said, it's still an existential threat to creative fields. It is scary, as you said. I think it'd be a great blow to humanity to have machines learn to "create" art as we do. Thus the no-AI rule exists on principle, even though AI stories are still not very good. It exists to promote human work, which I think is fair and shouldn't be changed.
 
JEsus H. Christ, I have a life.

I would like an "AI writer" or "AI written" category on Lit. That's it.

I am so sick with some people trying to argue with "anything I dislike is trolling, die troll" pseudo argument.
Fair enough. But we do get our fair share of people doing exactly what I described, around here, so it becomes a "if it walks like a duck" assumption. My apologies.
 
Never the less, I wanted to share, because it is an incredible story, which showcases what an unfiltered LLM can do.
Would you share a few paragraphs sample in this thread? I don't think that's breaking the rules.

I ask because most of the AI generated stuff I've seen is pretty yawn, but you seem jazzed about what it produced. It'd change my thinking on AI a bit if you're getting it to make stuff that impresses me.

Of course, you'll probably get some (aggressive) literary critique here. But you're kinda inviting that by making the thread?
 
Would you share a few paragraphs sample in this thread?


I asked for the story in a DM and AIWriter obliged. I found it very instructive. A lot of people here in AH seem to know what AI generated stories are like. I didn't. Now I know that at least one read like a grammar perfect encyclopedia of cliches. I was glad to find that out.
 
I asked for the story in a DM and AIWriter obliged. I found it very instructive. A lot of people here in AH seem to know what AI generated stories are like. I didn't. Now I know that at least one read like a grammar perfect encyclopedia of cliches. I was glad to find that out.
From what I've seen of AI-generated stories, I can imagine how a university professor feels when they have to grade first-year papers on Shakespeare's Sonnet no. 18. Line after line of trite banalities hung on hollow repetition of what the professor explained in class, showing the barest understanding of a few principles of literary analysis, with no insights or understanding of art or context or human nature, and concerned mostly with filling two pages by the deadline.
 
From what I've seen of AI-generated stories, I can imagine how a university professor feels when they have to grade first-year papers on Shakespeare's Sonnet no. 18. Line after line of trite banalities hung on hollow repetition of what the professor explained in class, showing the barest understanding of a few principles of literary analysis, with no insights or understanding of art or context or human nature, and concerned mostly with filling two pages by the deadline.
Excellent description (I assume it's accurate.) But how have you seen any AI stories?
 
People have posted snippets here in the forums. Not much lately, but there was a time last winter when there'd be one posted every week or so in Story Ideas or somewhere else.

There are also posts in reply to threads. A low post count account, and a paragraph of meaningless information that leaves you thinking, "Yeah, so?" And very often they end with something along the lines of "In conclusion, it's a complicated topic."

It's like the drinks dispenser in H2G2 that runs all kinds of analyses and then produces a drink that's almost, but not quite, entirely unlike tea.
 
I don't know if I should be happy it got rejected which means that despite the false rejections created by the no-AI rule, it does seem to be doing it's job, or mad that you even thought this was a good idea considering the issues the AI stuff has created over the past year.
 
From what I've seen of AI-generated stories, I can imagine how a university professor feels when they have to grade first-year papers on Shakespeare's Sonnet no. 18. Line after line of trite banalities hung on hollow repetition of what the professor explained in class, showing the barest understanding of a few principles of literary analysis, with no insights or understanding of art or context or human nature, and concerned mostly with filling two pages by the deadline.
To be fair, who could come up with anything new to say about Shakespeare at this point? Even the professor is probably retreading talking points that are a couple of centuries old. If the class responds by regurgitating received knowledge on command... well, they may have learned the underlying lesson after all. ;)
 
Yeah, 150+ stories every day is totally not enough; we clearly need the whole internet posting their AI-generated stories. When it comes to the quality of what you "wrote", no offense but I doubt it's really an incredible story, although it's probably not worse than the average here.
If there's one thing I support Laurel in, that's her firm stance on AI-generated junk. Here's hoping she tightens her standards with the human-written junk too. :p
If she increases her standards with human junk how’m I gonna get my stuff on here.

Cut me some slack, Set.
 
I don't think when this site started, they ever saw it getting to this point. 150 a day with a que backed up a few days is a lot.

I think NC stories would need more than a quick glance to determine if they're NC or DC/reluctance, but the site refuses to bring in any help even at a volunteer level. Meanwhile, I think anyone that can think on it unbiasedly should realize Mind Control is 100% non consensual with no debate possible. So, you have MC and a category called NC, then pretend to have an issue with someone coming here thinking "Do they allow rape fantasies? Oh, cool, Non consent, guess they do." The same way I came here long ago wondering if I could post incest and saw they have an incest category.

But incest is all fine, NC is a misnomer in context of 'rules' and don't tell me they're not capable of going in and changing a category title. "Reluctance/rough sex" works. On that note NC/R are two totally different things so having them together is a joke, and leads to this mess.

Again, I'll say I'm not kink shaming, want to write and read rape fantasies, go for it. Site wants them here? Sure, why not? The double talk-and endless support of said doubletalk by site bootlicks-is tiresome

Don't get me started on the infamous bi-sexual category bold faced lie.
When reading this thread I was gonna suggest some volunteer moderators to read submissions and help lessen the workload. I guess Laurel is against this for some reason?

I can't imagine being part of a two person team needing to approve 150 stories a day. There's just no way to fully read all of them. When I first started coming onto this site and saw they had a Nonconsent category, I thought that was for fantasy rape stories. It wasn't until I read the guidelines before submitting my own work that I realized I was mistaken. I imagine a lot of people don't even bother reading those guidelines before submitting.

To comment on the purpose of this thread, I definitely agree with the no-AI rule.
 
When reading this thread I was gonna suggest some volunteer moderators to read submissions and help lessen the workload. I guess Laurel is against this for some reason?

I can't imagine being part of a two person team needing to approve 150 stories a day. There's just no way to fully read all of them. When I first started coming onto this site and saw they had a Nonconsent category, I thought that was for fantasy rape stories. It wasn't until I read the guidelines before submitting my own work that I realized I was mistaken. I imagine a lot of people don't even bother reading those guidelines before submitting.

To comment on the purpose of this thread, I definitely agree with the no-AI rule.
If Laurel was ever going to consider opening up story moderation to volunteers, it would have happened by now. You open yourself to a cascade of potential problems once you start delegating and giving people access to what's behind the curtain, and it only takes one bad apple to cause a cataclysm. She's obviously weighed both sides of that equation, and decided that slogging away 8+ hours every single day doing it herself is less of a headache.
 
It was interesting to read what the author believes distinguishes her from the AI. Flying out to do research, getting authentic information, that kind of thing. Good for her, but it didn't add anything to the reading experience for me.

With the AI story, I lost interest nearly immediately. I skimmed to the end, but all the exposition at the beginning didn't seem to serve any purpose. And then when it comes to the nitty-gritty, the actual encounter, it was "They spoke about their lives, their interests and the little quirks that made them unique." Going back to the great "show v tell" debate, we're shown Lydia in her daily life, and then we're told she has an encounter that could shake it up. The whole thing could have been summarised in just that one sentence without losing much.
 
It was interesting to read what the author believes distinguishes her from the AI. Flying out to do research, getting authentic information, that kind of thing. Good for her, but it didn't add anything to the reading experience for me.

With the AI story, I lost interest nearly immediately. I skimmed to the end, but all the exposition at the beginning didn't seem to serve any purpose. And then when it comes to the nitty-gritty, the actual encounter, it was "They spoke about their lives, their interests and the little quirks that made them unique." Going back to the great "show v tell" debate, we're shown Lydia in her daily life, and then we're told she has an encounter that could shake it up. The whole thing could have been summarised in just that one sentence without losing much.

I believe that one sentence is called a "prompt" ;-)

The AI-written story is obvious there. I saw another one that had ten pieces of flash fiction, half written by humans and half written by GPT-4 to a prompt, and that was harder; I ended up getting only seven out of ten correct. But those were very short, only a few paragraphs, and that's where LLMs are strongest since it takes time for some of their deficiencies to become apparent.
 
Back
Top