How many typos do I need to add to not be flagged as AI?

[Just4Sheets]

I'm getting increasingly frustrated with Literotica's "possible AI" flag.

I do not use AI to write my stories. I also don't use it to rewrite or polish them. Every word is mine. Before I submit, I read each story out loud to my wife and mark any issues I catch that way.

Still, my stories keep getting sent back with the same generic "this might be AI" message.

Each time, I resubmit with a clear and polite note to the admins. So far, no progress. At this point, I have six stories stuck in pending, and a couple of those were originally submitted several months ago.

I've been a reader here for over 15 years and wanted to give back as a writer. But lately I'm starting to wonder if Literotica is still the right place for me.

If anyone else has dealt with this and found a solution, I'd really appreciate the advice. And if any admins happen to read this, I'd truly like to understand what I can do differently to avoid being flagged.


J4S

 

[AwkwardlySet]

Seven.

Jokes aside, you can see that Lit is becoming a huge mess. Maybe see if @AwkwardMD's thread can help you figure out the problem. Forget trying to appeal or even reach website admins. That's not going to happen.

Other than that, I can only suggest other story websites.

 

[BobbyBrandt]

You might want to try submitting differently.

If you typically paste your content into the submission field, try uploading it as a file instead.

If you typically submit by uploading a file, try pasting the content in the submission field instead.

 

[Just4Sheets]

You might want to try submitting differently.

If you typically paste your content into the submission field, try uploading it as a file instead.

If you typically submit by uploading a file, try pasting the content in the submission field instead.

Seems weird, but I'm willing to try when I resubmit.

Thanks!


J4S

 

[BobbyBrandt]

Seems weird, but I'm willing to try when I resubmit.

Thanks!


J4S

Please share your results.

 

[SmilingLez]

I'm getting increasingly frustrated with Literotica's "possible AI" flag.

I do not use AI to write my stories. I also don't use it to rewrite or polish them. Every word is mine. Before I submit, I read each story out loud to my wife and mark any issues I catch that way.

Still, my stories keep getting sent back with the same generic "this might be AI" message.

Each time, I resubmit with a clear and polite note to the admins. So far, no progress. At this point, I have six stories stuck in pending, and a couple of those were originally submitted several months ago.

I've been a reader here for over 15 years and wanted to give back as a writer. But lately I'm starting to wonder if Literotica is still the right place for me.

If anyone else has dealt with this and found a solution, I'd really appreciate the advice. And if any admins happen to read this, I'd truly like to understand what I can do differently to avoid being flagged.


J4S

I'm beginning to think that the AI message is also being generated for any work that sits in the pending queue more than 3 weeks in order to purge the database. I believe it's a catch-all message versus an actual flagging done by an AI tool to catch AI work. YMMV

 

[Just4Sheets]

Please share your results.

I have stories resubmitted 11/30 and 12/1, so I'll wait for those to hit 15 days and then resubmit all 6.

J4S

 

[Just4Sheets]

I'm beginning to think that the AI message is also being generated for any work that sits in the pending queue more than 3 weeks in order to purge the database. I believe it's a catch-all message versus an actual flagging done by an AI tool to catch AI work. YMMV

That would make as much sense as any other reason!

J4S

 

[iwatchus]

It really isn't a lack of mistakes that makes something look like AI. Can people get that out of their head?

I don't know what they are using to check for AI (and couldn't say if I did). I'm absolutely positive it's not counting typos in any way.
And I strongly believe that they are not using one of the worthless on-line AI checkers.
Unfortunately, no AI checker (human or automated and the belief is they are using a combination of both) is perfect. And there will be some false positives and some false negatives.

Since you state you did not use AI, I will presume you are a false positive. Which really sucks. I;m sorry you are going through this.

@AwkwardMD any help for him?

 

[EmilyMiller]

This area is rapidly descending into superstition and folklore. Maybe always wear clean socks when you submit your story. It might help.

We have a chaotic situation and many are seeking patterns and explanations that may not exist.

 

[ShelbyDawn57]

I don't recommend adding grammatical errors, but I'd suggest you look at things like varying your sentence and paragraph length, adding colloquialisms, Vary your dialog tags; don't just say 'he said, she said' all the time, or maybe us it more. AI is predictable, follows patterns, uses the same words over and over again. Change things up a bit. AI isn't creative at all. It just regurgitates what's in its model. You're writing stories, that makes you better than AI out of the gate. Show it in your words.

And for those 'style' purists out there. Based on the research I've read*, AI is defined by traits and characteristics of style. If you're getting flagged, your style is too similar to that of an AI, well, that's your call.

*I did a Google search for AI flagging in fiction and read the three or four accredited(not Wikipedia) articles I saw and that all seemed to say the same thing.

 

[Just4Sheets]

I'm guessing my neurodivergence is playing a role.

Not much I can do about that.

Maybe I'll get an editor .. I don't know.


J4S

 

[LisaMayTurner]

There are so many stories that use the same patterns of over-used porn language and the "same old stuff" that gets recycled and re-re-plagiarised. Such things are for more likely to trigger Ai concerns as Ai function to gather up all the usual "same old stuff" and regurgitate it.

 

[EmilyMiller]

I'm guessing my neurodivergence is playing a role.

I’m autistic. I’ve never been flagged for AI. Again this is getting into forcing square explanations into round problems.

 

[Just4Sheets]

I’m autistic. I’ve never been flagged for AI. Again this is getting into forcing square explanations into round problems.

As you know, it's a spectrum. Autism affects all of us in different ways. My daughter can't have a real conversation with anyone. My son tries but winds up being shunned as "creepy." I was just weird.

I just went to openai and pasted one of my stories and it said it was likely AI. I'm not sure what to do. Maybe it's best if I just take a break.


J4S

 

[ShelbyDawn57]

As you know, it's a spectrum. Autism affects all of us in different ways. My daughter can't have a real conversation with anyone. My son tries but winds up being shunned as "creepy." I was just weird.

I just went to openai and pasted one of my stories and it said it was likely AI. I'm not sure what to do. Maybe it's best if I just take a break.


J4S

Don't be too hard on yourself. I just read Haven't You Ever Wondered. It's a good story, creative and well written. I'd suggest trying to be more descriptive and drawing things out a bit, but otherwise, it has me wanting to read more of you work. It also proves you can write in a way that doesn't get flagged. Maybe compare your new stuff to the style you used in Wondered, and you'll see some things.

Good luck.

 

[EmilyMiller]

As you know, it's a spectrum. Autism affects all of us in different ways. My daughter can't have a real conversation with anyone. My son tries but winds up being shunned as "creepy." I was just weird.

It’s not just a spectrum - it’s a series of axes with any given individual having a different set of coordinates. It’s no where near as simple as even the word salad of acronyms we use to describe those of us who are outliers on some of those axes. Some people with ASD need around the clock support. Some - like me - have post grad qualifications and manage teams. There is no unitary ND experience.

I just went to openai and pasted one of my stories and it said it was likely AI. I'm not sure what to do. Maybe it's best if I just take a break.

They don’t work (anymore than the Literotica one does).

 

[PennyThompson]

I just went to openai and pasted one of my stories and it said it was likely AI.

Also not a reliable way to check for AI, ChatGPT is not designed or trained to do that kind of language analysis, and will literally just make things up in order to respond to a user's prompt.

AI-powered AI detectors aren't particularly reliable, and ChatGPT isn't even that

 

[EmilyMiller]

Also not a reliable way to check for AI, ChatGPT is not designed or trained to do that kind of language analysis, and will literally just make things up in order to respond to a user's prompt.

AI-powered AI detectors aren't particularly reliable, and ChatGPT isn't even that

This x 1000

GenAI lies when it can’t find a statistical match. It also makes mistakes frequently when it can. Eventualy there will be serious screw ups and we will wonder why we worshiped at this empty temple.

 

[SmilingLez]

There are so many stories that use the same patterns of over-used porn language and the "same old stuff" that gets recycled and re-re-plagiarised. Such things are for more likely to trigger Ai concerns as Ai function to gather up all the usual "same old stuff" and regurgitate it.

Unfortunately most stories / videos dealing with sexuality are rehashed themes. I am sure most of us can figure out the plot of a Literotica story after reading the first few paragraphs.

 

[Just4Sheets]

Don't be too hard on yourself. I just read Haven't You Ever Wondered. It's a good story, creative and well written. I'd suggest trying to be more descriptive and drawing things out a bit, but otherwise, it has me wanting to read more of you work. It also proves you can write in a way that doesn't get flagged. Maybe compare your new stuff to the style you used in Wondered, and you'll see some things.

Good luck.

Thank you!

I feel like some of my pending stories are similar, but some are very different.

I appreciate you reading and am glad you enjoyed the story.


J4S

 

[PennyThompson]

This x 1000

GenAI lies when it can’t find a statistical match. It also makes mistakes frequently when it can. Eventualy there will be serious screw ups and we will wonder why we worshiped at this empty temple.

See my amateurish data analysis of comments on a Loving Wives story, out of curiosity I tried using AI to help with the data entry and qual coding portion of the project, and it was completely helpless

It made up numbers and confidently miscounted, contradicted itself from one prompt to the next, repeatedly claimed to have capabilities it didn't actually have and then "apologized" when I asked it to take an action that it said it was able to do but wasn't...

Fundamentally, LLMs are constitutionally unable to judge the accuracy of their own output or differentiate between hallucination and fact. They just can't, it's not a function that they have or will likely ever have.

 

[VerbalAbuse]

I can only suggest other story websites.

please do.

 

[StillStunned]

It made up numbers and confidently miscounted, contradicted itself from one prompt to the next, repeatedly claimed to have capabilities it didn't actually have and then "apologized" when I asked it to take an action that it said it was able to do but wasn't...

I had a colleague like that once.

 

[Lifestyle66]

Big Brain says the answer is "42".