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

As a reader, I hope that you don't give up

1765737007197.png

Thanks! I'm hoping my stories get published. I tried with the one story so far. I had 6 pending and pulled them all back .. and then worked on one and submitted it. (Note: The other 2 drafts aren't in any condition to be published)

We'll see what happens.


J4S
 
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.​

They don’t work (anymore than the Literotica one does).
Understanding the complexity, I am surprised you made the dismissive comment that your did:

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


At one point in this space it might not lead you to write in a manner likely to be inaccurately flagged. That says nothing about how other points, regions in this space might effect your writing.​

Unless you disbelieve the OP, the flagging system is inaccurate. I would ask that focus be on help them navigate through this imperfect system​
 
No AI checkers really work, even semi well. People who are not neurodivergent get pinged for AI and people like me who are ND don’t. It’s not about the writer, it’s about the fucking awful tool the site users (and they are all pretty fucking awful).
 
Understanding the complexity, I am surprised you made the dismissive comment that your did:​
I’m autistic. I’ve never been flagged for AI. Again this is getting into forcing square explanations into round problems.
At one point in this space it might not lead you to write in a manner likely to be inaccurately flagged. That says nothing about how other points, regions in this space might effect your writing.​
Unless you disbelieve the OP, the flagging system is inaccurate. I would ask that focus be on help them navigate through this imperfect system​

I don't blame anyone who doesn't believe me. It's a hard sell. I linked my story on Reddit so people could decide for themselves. I don't believe any tools are at the level they could write a story like that. I have no idea why my stories got sent back and I don't think I ever will. But here's how it went.

I was thinking about my son, who is a college student living on campus. I was thinking about the world now compared to how it was when I was his age. And I came up with the thought, "if I was his age and single in this day and age, and I met and was attracted to a woman who wound up having a penis, how would I handle that?" And I started writing a story.

Like all my stories I had no idea where it was going. I wrote the entire thing as a hybrid outline/story. Then I added more meat. Next was the dialog, because I hate writing dialog. And finally I filled in the explicit sex (prior to this step the story just said things like "[Fellatio Here]". To do that last part I read a bunch of other stories because I don't know what many of those sex acts actually feel like.

Finally I read the story out loud to my wife and noted anything that just felt wrong while I was reading it - that's my grammar checker. She liked it, so I published it and it got bounced back at possible AI.

I'd have been surprised, but it was far from the first time. i added an admin note that essentially said, "This isn't AI. Please actually read it and publish it. Your AI detector is making a mistake." And that was about 3 weeks ago.


J4S
 
I don't believe any tools are at the level they could write a story like that.
The detection tools are as fallible AF. I have spoken to so many people who have got caught. Including ones where I’ve been privy to the [non-AI] writing process. I’ve even messaged the site on their behalf.

I have sympathy with Literotica (for some things). What are they meant to do? The actual problem lies with those who are undoubtedly using AI to avoid the effort of writing.
 
I have sympathy with Literotica (for some things). What are they meant to do? The actual problem lies with those who are undoubtedly using AI to avoid the effort of writing.

Oh, I get it .. I truly do! They have an impossible job screening all stories for AI. But maybe it's not so impossible. Maybe they don't need to pre-screen at whatever level the detection tool is set. Maybe they can screen for only the most blatant cases and add a way for readers to report a story as "AI Slop." And then just those stories can be gone over with more scrutiny.

And make a way for less targeted reporting. Make it at least a little difficult for the "Loving Wives Hate Club" to just report everything where the woman wins. Reporting done by real accounts only. Accounts with Reports that don't result in AI detection (so invalid or incorrect votes) aren't considered as strongly as others. So a certain number of votes for AI Slop equals a full review of the story. If the full review doesn't agree it's AI then those 10 voters' votes are now each 0.9 votes each. And the next time .. 0.8 votes.

If the story, after review, is considered AI punish the author. Don't let them submit stories. Give them a single warning and then no submitting. Do literally anything else! If after scrutiny the story is considered to be AI I don't care what they do to the author's account.

This isn't my area, so I'm spit-balling here. But there has to be a better way than just having some fallible engine decide my hard work seems suspect and then .. I'm just screwed.


J4S
 
Oh, I get it .. I truly do! They have an impossible job screening all stories for AI. But maybe it's not so impossible. Maybe they don't need to pre-screen at whatever level the detection tool is set. Maybe they can screen for only the most blatant cases and add a way for readers to report a story as "AI Slop." And then just those stories can be gone over with more scrutiny.

And make a way for less targeted reporting. Make it at least a little difficult for the "Loving Wives Hate Club" to just report everything where the woman wins. Reporting done by real accounts only. Accounts with Reports that don't result in AI detection (so invalid or incorrect votes) aren't considered as strongly as others. So a certain number of votes for AI Slop equals a full review of the story. If the full review doesn't agree it's AI then those 10 voters' votes are now each 0.9 votes each. And the next time .. 0.8 votes.

If the story, after review, is considered AI punish the author. Don't let them submit stories. Give them a single warning and then no submitting. Do literally anything else! If after scrutiny the story is considered to be AI I don't care what they do to the author's account.

This isn't my area, so I'm spit-balling here. But there has to be a better way than just having some fallible engine decide my hard work seems suspect and then .. I'm just screwed.


J4S
I think your assumption here is that the AI checkers work and can be calibrated. Whatever setting the site adopts there will be false positives and false negatives.

From convos with the site earlier in the year, the volume of AI written submissions was becoming problematic back then. I can only assume it has gotten worse.

The sooner the AI bubble bursts the better. It will trash the real economy too, but the longer the bubble grows, the bigger the bang when it inevitably pops.

The AI boom isn’t to do with actual companies deriving benefits. It’s a Ponzi scheme involving chip makers, data center fabricators, tech behemoths, and AI companies all paying each other to buy or use their products. There are no profits, just horrendous debt often creatively off balance sheet. And the debt is being sliced up just like sub-prime mortgages were before. Humans always choose greed, we are such a stupid species.
 
From convos with the site earlier in the year, the volume of AI written submissions was becoming problematic back then. I can only assume it has gotten worse.

The sooner the AI bubble bursts the better. It will trash the real economy too, but the longer the bubble grows, the bigger the bang when it inevitably pops.

The AI boom isn’t to do with actual companies deriving benefits. It’s a Ponzi scheme involving chip makers, data center fabricators, tech behemoths, and AI companies all paying each other to buy or use their products. There are no profits, just horrendous debt often creatively off balance sheet. And the debt is being sliced up just like sub-prime mortgages were before. Humans always choose greed, we are such a stupid species.
I don't know much about AI. My daughter sends me comics she makes using it. I see videos pop up on facebook where I have to say, "well, that's not real." But I don't really get it.

I haven't been here so long. I could give the story about how at my age I suddenly decided to start writing erotic fiction, but it's not that exciting. The easiest thing I could say is that a year ago I would have never even considered this stuff. And once I started .. I was addicted. Checking for comments every day. Looking to see when my story got published. And now, I'm in withdrawal .. and the longer it goes on, the more I think maybe I'm over the addiction. And that makes me sad.


J4S

PS: I just saw you have a link for your "Rollo Tomasi" stories. Just wanted to say, "I understood that reference" 😉
 
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

It’s not about typos. It’s about style, sentence structure, cadence and repetition. I’ve read some of your stories, and it’s very obvious you’re using AI. I don’t personally care, and anyone who does needs to just get over themselves, because AI is here to stay.

It's like when people freaked out after cars were first invented. Cities literally banned them. And what happened? We all drive cars now. But until we get to the point where AI has been around long enough that society stops clutching its pearls, you’re going to have to be smarter than the GPT, and right now, you’re not.

Start paying attention to the words it repeats constantly and swap them out. Words like: sharp, steady, deliberate, stepped. These words show up all the time in AI-produced creative writing. One of your 3.7k-word stories repeats “steady” 7 times alone.

AI also has a habit of writing in threes and repeating sentence openings. Here’s dialogue from one of your stories:

“Not in theory. Not in fantasy. In reality.”

Nobody talks like that. A real person would say, “Not in theory or fantasy, but in reality.”

The same pattern shows up constantly:

“Gray walls, no windows, nothing left out.”

“The eyes, maybe, but the jaw was softer, cheekbones higher, hair spilling in uneven waves across bare shoulders.”

“My lungs felt smaller, the rhythm faster, every breath pushing differently against my ribs.”

You starting to get it yet? That rhythm is a dead giveaway. You can use AI, but if you want to pass for human, you’ve got to rewrite those lines. Otherwise, you’re not fooling anyone except yourself.
 
It’s not about typos. It’s about style, sentence structure, cadence and repetition. I’ve read some of your stories, and it’s very obvious you’re using AI. I don’t personally care, and anyone who does needs to just get over themselves, because AI is here to stay.

It's like when people freaked out after cars were first invented. Cities literally banned them. And what happened? We all drive cars now. But until we get to the point where AI has been around long enough that society stops clutching its pearls, you’re going to have to be smarter than the GPT, and right now, you’re not.

Start paying attention to the words it repeats constantly and swap them out. Words like: sharp, steady, deliberate, stepped. These words show up all the time in AI-produced creative writing. One of your 3.7k-word stories repeats “steady” 7 times alone.

AI also has a habit of writing in threes and repeating sentence openings. Here’s dialogue from one of your stories:

“Not in theory. Not in fantasy. In reality.”

Nobody talks like that. A real person would say, “Not in theory or fantasy, but in reality.”

The same pattern shows up constantly:

“Gray walls, no windows, nothing left out.”

“The eyes, maybe, but the jaw was softer, cheekbones higher, hair spilling in uneven waves across bare shoulders.”

“My lungs felt smaller, the rhythm faster, every breath pushing differently against my ribs.”

You starting to get it yet? That rhythm is a dead giveaway. You can use AI, but if you want to pass for human, you’ve got to rewrite those lines. Otherwise, you’re not fooling anyone except yourself.
IMG_3004.jpeg

Dude! This your very first post and you choose to make it about this? What did you do create an alt as you didn’t want the blowback? Or are you just playing the provocative card that is getting more and more common around here?

Oh, and welcome. And I look forward to you posting something constructive.
 
It’s not about typos. It’s about style, sentence structure, cadence and repetition. I’ve read some of your stories, and it’s very obvious you’re using AI. I don’t personally care, and anyone who does needs to just get over themselves, because AI is here to stay.
It's hard to respond to your post, because the obvious thing is to say, "You're wrong," and then I just sound defensive. Instead I'll just go with .. Good to know.

I wonder, assuming you're correct, where AI could have gotten that "rhythm"? I've heard it learns by reading what already exists. So my guess is that what you say is obvious AI is something it learned by examining non-AI literature.

But yeah .. you gave examples and you're correct that they all share a certain "cadence."

Maybe in my next story I'll avoid that "writing in threes" thing. Maybe I'll avoid the word "steady." I wonder if that will make my story better or worse. Either way the story won't be AI.


J4S
 
But I'm a fairy,
and so when able,
I must follow the rule of three.

I think the writing in three thing comes from very old stories, especially ones with fairies in them.
 
But I'm a fairy,
and so when able,
I must follow the rule of three.

I think the writing in three thing comes from very old stories, especially ones with fairies in them.
I think many of these ‘AI does X’ comments are BS. What it does often seem to do is to have an unvaried, homogenous, repetitive style, so much so that it becomes bland. But the ‘don’t use em dashes and avoid AI detection’ advice seems off.

It’s one reason I get upset by the ‘people with ASD write like AI’ claims. It’s the fucking Sheldon Cooper crap that we are all unemotional robots with zero understanding of our fellow humans and zero empathy.

I call bullshit on that.
 
It’s one reason I get upset by the ‘people with ASD write like AI’ claims. It’s the fucking Sheldon Cooper crap that we are all unemotional robots with zero understanding of our fellow humans and zero empathy.

I'm sorry, since I know that's what I said. I never thought I wrote like AI before my stuff got sent back here.

For now I'm going to sit back and wait for the one story I have pending right now to be Published or Sent Back. Until then I'm not sure how to proceed.

Fingers crossed.


J4S
 
I think many of these ‘AI does X’ comments are BS. What it does often seem to do is to have an unvaried, homogenous, repetitive style, so much so that it becomes bland. But the ‘don’t use em dashes and avoid AI detection’ advice seems off.

It’s one reason I get upset by the ‘people with ASD write like AI’ claims. It’s the fucking Sheldon Cooper crap that we are all unemotional robots with zero understanding of our fellow humans and zero empathy.

I call bullshit on that.
Was discussing just this with my autistic son(high functioning Asperger's). His opinion is Sheldon isn't autistic, he's just an asshole.
 
But I'm a fairy,
and so when able,
I must follow the rule of three.

I think the writing in three thing comes from very old stories, especially ones with fairies in them.
Maybe there's something to that. There's almost certainly something to it, in terms of why threes can feel cliché.

Threes can also a very satisfying rhetorical device. I think it's neurological. Of course someone can fuck it up but I don't think the reason for threes being commonly employed has everything to do with past conventions. I believe it has more to do with how effective they can be.
 
Last edited:
Back
Top