The AI Rejection Conversation Matters

Does anyone know if there are any reliable AI detection that you can you to figure out what part(s) of your story is triggering the AI filter? Because I am at a loss as to why my stories have been rejected.
No, nothing at all. The management isn't telling what is triggering and everyone is just guessing at ways to edit and rewrite.
 
Does anyone know if there are any reliable AI detection that you can you to figure out what part(s) of your story is triggering the AI filter? Because I am at a loss as to why my stories have been rejected.
There are no reliable AI detectors. That's the (main) problem. Even the companies that make AI detectors are like, "hey, so maybe don't use these for anything important? Kthxbai." The only thing you could do is to use the same detector that Literotica is using, because it will (probably) give the same results multiple times. But if the site gives out which detector they use, then people will use that detector to skirt the detector, and...
 
There are no reliable AI detectors. That's the (main) problem. Even the companies that make AI detectors are like, "hey, so maybe don't use these for anything important? Kthxbai." The only thing you could do is to use the same detector that Literotica is using, because it will (probably) give the same results multiple times. But if the site gives out which detector they use, then people will use that detector to skirt the detector, and...

I know that AI detectors are only looking for specific patterns, and that the writing of some people unfortunately matches those patterns. I was just wondering if anyone had figured out if one service or another was better at detecting the patterns that Literotica's detector detects.
 
  1. There are not wholesale AI-related rejections - more stories are published here than ever
  2. 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!
  3. There are edge cases, I’ve worked with some of them (maybe 8 now) my observation was that their natural style (giving the benefit of the doubt) was AI-adjacent and their work read a bit uncanny valley

I don't believe any of this for one second ....

I can think hundreds of stories falsely flagged for AI
 
I don't believe any of this for one second ....

I can think hundreds of stories falsely flagged for AI
Given that thousands of stories get posted a month, hundreds over the course of, say, a year and a half, is a relatively small percentage.

ETA: an "a OR the" search, limited to the last year, returns 59216 stories, or about 5K stories a month. Even if 500 stories got rejected for AI usage this year alone, that's about 1%; not a small amount, but not huge, either, and well within the realm of possibilities, especially now that there's a GPT3.5-equivalent that can be run on a local machine.
 
Last edited:
Given that thousands of stories get posted a month, hundreds over the course of, say, a year and a half, is a relatively small percentage.

I'll believe both figures it when I see real accurate statistics .... if the site owners want to communicate anything directly backed up with evidence then I will consider it.

But so far I am yet to hear one single utterance on this whole issue ...yet I know for a fact that very talented established authors have left this site due to many false AI flags .... these are people that have been writing for 20+ years on this site and in one case someone who was writing on Rusty BBS in 1991.

What has/is happening to Literotica is a disgrace. Using a undisclosed, inaccurate AI tool to accuse long time authors of AI usage is frankly unconscionable especially since you have been living free of their writings for decades.
 
Last edited:
I'm curious what's triggered this latest spate of posts about AI rejections. It was largely quiet on that front for months, now all of a sudden people are complaining about it happening to them. Or people they know, at least.
 
I'll believe both figures it when I see real accurate statistics .... if the site owners want to communicate anything directly backed up with evidence then I will consider it.

But so far I am yet to hear one single utterance on this whole issue ...yet I know for a fact that very talented established authors have left this site due to many false AI flags .... these are people that have been writing for 20+ years on this site and in one case someone who was writing on Rusty BBS in 1991.

What has/is happening to Literotica is a disgrace. Using a undisclosed, inaccurate AI tool to accuse long time authors of AI usage is frankly unconscionable especially since you have been living free of their writings for decades.

I literally gave you the statistics for how many stories were published in the last year. It's user-accessible data.

Look, I don't disagree at all that the AI "detection" stuff is problematic at best and absolutely nonsense at worst. However, a ton of writers are still making it through the gauntlet. Should there be some leeway for longstanding contributors? Yes. In fact, I'd guess there probably is, at least in some cases.

Given that a particularly vigorous pro-AI detection schmuck plugged all of my stories into one of the detectors and reported it to the admins (which went absolutely nowhere), I'm sympathetic to the folks worried about witchhunting/mass censorship. However, I'm also sympathetic to the site trying to stop AI-written stories from getting published, and, given human nature, the idea of even established authors falling to the dark side is a very real possibility. Right now, there are no great answers, and we need to be cognizant of that.
 
I'm curious what's triggered this latest spate of posts about AI rejections. It was largely quiet on that front for months, now all of a sudden people are complaining about it happening to them. Or people they know, at least.
Literotica stopped paying dividends.


:p
 
Literotica stopped paying dividends.


:p
I'd love to be a fly on the wall at that shareholders meeting.

"The Company has put a moratorium on dividend distributions. Management believes that the shareholders have been squeezing, milking, sucking (soft moan) the Company dry. The thrust (whimper) of our policy is that we will continue to plough on vigorously (moan), doing all the hard, hard work (pant) until we are satisfied." (Collapses in a happy heap and lights a cigarette).
 
I'd love to be a fly on the wall at that shareholders meeting.

"The Company has put a moratorium on dividend distributions. Management believes that the shareholders have been squeezing, milking, sucking (soft moan) the Company dry. The thrust (whimper) of our policy is that we will continue to plough on vigorously (moan), doing all the hard, hard work (pant) until we are satisfied." (Collapses in a happy heap and lights a cigarette).

I feel dirty now.
 
I literally gave you the statistics for how many stories were published in the last year. It's user-accessible data.
I literally don’t believe anything that is not independently audited… unless you have counted every single published story and checked it for being correct and referenced that against every single story submission then I don’t believe anything that this sites admins purportedly put out as data.
 
I literally don’t believe anything that is not independently audited… unless you have counted every single published story and checked it for being correct and referenced that against every single story submission then I don’t believe anything that this sites admins purportedly put out as data.
Ah, so now we’ve wandered into the realms of paranoia.

Look, go to the search page, type in “a OR the” (minus the quotes), change the settings to only return results from the last year, and hit the button. That should return every work published in the last year that includes either “a” or “the” somewhere in the title or text; if that’s not every single title published, it’s certainly close enough to meet the nebulous standards established by your “I know of hundreds of stories that were rejected.”

Oh, you do? Give us a list, then.

But, of course, you can’t, because you said something hyperbolic without the ability to back it up. Then, instead of admitting that, you chose to set impossible goals for the people participating in the discussion. And I’ll bet that, even if Laurel and Manu coughed that data up, you’d then say, “I want direct access to the database.” And then, when that failed to satisfy, “I don’t believe this is the real database.”

Tens of thousands of stories get published on the site every year. Some will get rejected for spurious purposes; mine did, twice, both for non-AI-related reasons. A simple resubmission with a message will usually get a returned story through the process. When it doesn’t… them’s the breaks. Complain about it on the forums, go elsewhere, or chalk it up to dumb luck, but don’t expect some ridiculous pie in the sky resolution like “I want to audit all of your data!” It’s a private site. They can literally reject any submission for any reason up to and including “because I said so.” Deal with it or move on.
 
Last edited:
Ah, so now we’ve wandered into the realms of paranoia.

Look, go to the search page, type in “a OR the” (minus the quotes), change the settings to only return results from the last year, and hit the button. That should return every work published in the last year that includes either “a” or “the” somewhere in the title or text; if that’s not every single title published, it’s certainly close enough to meet the nebulous standards established by your “I know of hundreds of stories that were rejected.”

Oh, you do? Give us a list, then.

But, of course, you can’t, because you said something hyperbolic without the ability to back it up. Then, instead of admitting that, you chose to set impossible goals for the people participating in the discussion. And I’ll bet that, even if Laurel and Manu coughed that data up, you’d then say, “I want direct access to the database.” And then, when that failed to satisfy, “I don’t believe this is the real database.”

Tens of thousands of stories get published on the site every year. Some will get rejected for spurious purposes; mine did, twice, both for non-AI-related reasons. A simple resubmission with a message will usually set a returned story through the process. When it doesn’t… them’s the breaks. Complain about it on the forums, go elsewhere, or chalk it up to dumb luck, but don’t expect some ridiculous pie in the sky resolution like “I want to audit all of your data!” It’s a private sure. They can literally reject any submission up to and including for any reason up to and including “because I said so.” Deal with it or move on.
comfort-comfy.gif
 
Look, go to the search page, type in “a OR the” (minus the quotes), change the settings to only return results from the last year, and hit the button. That should return every work published in the last year that includes either “a” or “the” somewhere in the title or text; if that’s not every single title published, it’s certainly close enough to meet the nebulous standards established by your “I know of hundreds of stories that were rejected.”
Plus the 25 or so stories of mine that have neither "a" nor "the".
 
For the record: according to the story tags portal, the site is currently hosting 622,015 stories. On Oct 16 a year ago, the total was 577,977. That's about 44k stories added in a year... or the net amount, anyway, since we can assume some were removed by the site or the author during that span.
 
Back
Top