AI Allegations Thread

[Bamagan]

https://forum.literotica.com/thread...thread-or-is-it.1593945/page-11#post-97454832

That's about where things went off the rails on that thread. Her homie is called Thavipah, but none of their content is searchable unless you read the replies that quote him/her/it.

Enjoy!

New business idea... create a searchable database of all the 'best' message board conversations on the internet and name it Coprolite.com... although someone apparently owns that domain already and redirects traffic.

 

[alohadave]

we had a brief handbagging over it.

Handbagging?

 

[onehitwanda]

Handbagging?

Handbags at dawn. The middle-aged lady of the world's alternative to a Walk-off.

 

[Bamagan]

Handbagging?

https://media4.giphy.com/media/ABgNxvbk95V3W/giphy.gif

 

[onehitwanda]

https://media4.giphy.com/media/ABgNxvbk95V3W/giphy.gif

I was just looking at this exact same gif.

My money's on Blue Granny.

 

[Bamagan]

I was just looking at this exact same gif.

My money's on Blue Granny.

She definitely seems to have more zap in her slap.

 

[onehitwanda]

She definitely seems to have more zap in her slap.

Probably has a bottle of Baileys and a half-brick in there. Baileys for getting it moving when it should but isn't, half-brick for stopping it moving when it is but shouldn't.

 

[Zootonius]

I was just looking at this exact same gif.

My money's on Blue Granny.

And for those old enough to remember Laugh-In
https://66.media.tumblr.com/98dfc8f242ff340f972b0f5855dec30d/tumblr_min72trIif1ram4r5o1_500.gif

 

[theWollstonecraftWoman]

A wee boot concerned how long it is taking to get my new tail up. Of course, it's been up for sale for several months. I don't think there is an issue with AI on it; I think I just submitted it on a Thursday, and the powers that be, Laurel, hasn't had a chance to authorize it. Crossing fingers, it goes up soon.

 

[MillieDynamite]

Sweetie, its tale, not tail.

A wee boot concerned how long it is taking to get my new tail up. Of course, it's been up for sale for several months. I don't think there is an issue with AI on it; I think I just submitted it on a Thursday, and the powers that be, Laurel, hasn't had a chance to authorize it. Crossing fingers, it goes up soon.

 

[alohadave]

A wee boot concerned how long it is taking to get my new tail up. Of course, it's been up for sale for several months. I don't think there is an issue with AI on it; I think I just submitted it on a Thursday, and the powers that be, Laurel, hasn't had a chance to authorize it. Crossing fingers, it goes up soon.

My latest has been pending for a week and it's nerve wracking when you are used to quick approvals.

I keep going back and forth between being worried that it's flagged for AI, she's just really busy, and it's stuck in Limbo.

 

[oneagainst]

I'm in the process of running up a new story arc featuring an AI character. I'm going to use AI tools to generate segments, particularly the AI's monologue to the FMC, as art for art's sake. I want to make a pretty specific point about where I think this is all going, and rather than discussing in AH, the best place to do this is in a story.

There is no fate but what we make.

 

[oneagainst]

that sounds like an interesting story and I would read it. Hope the site doesn't flag it for ai because then you will probably have trouble posting anything else here

That would be the ultimate irony, that a story based on social comment of the threat of AI to our "business" gets blocked by the very same AI tools it's commenting about.

I guess the sex scenes need to be extra unique then to avoid the censor...!

 

[MillieDynamite]

Mary's out of pocket today. She submitted it late Wednesday night of last week. That made it, for all particle purposes, a Thursday submission. The sixth was the last day for submitting stories for the Winter Holiday Contest. All of those will reviewed and published before anything unrelated will be looked at. I don't know if she'll be kicked or not. I've read her work, and it doesn't read like AI. That said, we don't know what AI looks like to Laurel. I don't believe she has anything to worry about. The speed at which stories get published depends on so many variables, beyond AI, that sometimes it takes a while.

I submitted a story on Thursday and it posted on Saturday but maybe that has to do with the category more than anything else. If you see your submission date reset that is a pretty good indication you have been flagged for ai and it will get kicked back soon

 

[MillieDynamite]

AI, at present, reads inherently like lousy writing. This makes it challenging for new writers. We all have AI moments in stories, clunky sentences, glaring mistakes, or odd word usage. But it's usually not enough to flag it at more than a few percentage points on a checker. But that brings us to the question of the hour. How high a percentage is too high? No one is telling us that. If it is 90% certain that it is not AI, is that 10% too much for Laurel?

I hope she doesn't have anything to worry about either. To be fair I have not come across any writing here that was flagged for ai that actually reads like a bot wrote it. There is another thread where people shared parts of their rejected stories and while some folks piped up about how bad the writing was-and again I am proof the site sets a low bar for writing quality so that should not even be a factor lol-nobody said the stories read like ai. It is clear there are plenty of false positives already

 

[SyleusSnow]

It's insane we're reduced to this, but it looks like the market for evading these crap AI writing detectors is booming:

10 Best undetectable AI writing tools to bypass AI content detectors

 

[alohadave]

It's insane we're reduced to this, but it looks like the market for evading these crap AI writing detectors is booming:

10 Best undetectable AI writing tools to bypass AI content detectors

Build a better mouse trap and someone will build a better mouse.

(My iPhone keeps flagging random words as incorrect before deciding they are. It's freaking me out. I had to look up mouse to make sure it was correct, and it still looked wrong. I think the AI in autocorrect is messing with me.)

 

[MillieDynamite]

I used ChatGPT for the following. I consider this dreadful.

Prompt
Write a two paragraph introducing a happy woman to the reader.

ChatGPT
Meet Emily, a vibrant and joyful soul who radiates happiness from the moment you encounter her. Her infectious laughter can be heard echoing through any room, instantly lifting spirits and creating an atmosphere of warmth. With a perpetual smile adorning her face, Emily has an uncanny ability to find joy in even the simplest of moments, turning mundane tasks into delightful adventures.

Emily's positive energy is not only reflected in her beaming countenance but also in the way she engages with others. A natural optimist, she possesses a magnetic charm that draws people towards her. Whether she's sharing anecdotes of her latest escapades or simply offering a friendly greeting, Emily has an innate talent for making everyone around her feel genuinely happy. Her presence is a reminder that happiness is not just a state of mind for her; it's a way of life that she effortlessly shares with those lucky enough to be in her orbit.

Both Grammarly and ProWritingAid agree with me. Does any writer want their writing to sound like this reads

 

[MillieDynamite]

I ran some of the examples here through checkers, and there was a lot of red in those samples (red being the highest probability of being AI)—quite a lot of yellow and orange. I'm not saying they are using AI, but if it triggers a higher than 50% probability, it would likely be rejected here. But I don't know what the cut-off is. It would behove the writer to run their work through an AI checker and reduce the flagged sentences.

ok my point is people who do not write like this at all are getting wrongly rejected here. Hopefully the site will see your example here too and figure out that they are rejecting stuff written by humans not bots

 

[NoTalentHack]

It's insane we're reduced to this, but it looks like the market for evading these crap AI writing detectors is booming:

10 Best undetectable AI writing tools to bypass AI content detectors

Bonus fun: most of the companies that provide AI detectors also make tools to help evade AI detectors.

 

[NoTalentHack]

I used ChatGPT for the following. I consider this dreadful.



View attachment 2295762
Both Grammarly and ProWritingAid agree with me. Does any writer want their writing to sound like this reads

See, I don't consider this dreadful. I just don't consider it good. I could EASILY see this as part of someone's first story; the writing in After the Future is Gone is only marginally better, if I'm being 100% honest. I think that's the problem: AI detectors don't detect bad writing; they detect common writing: technically competent (in terms of grammar, spelling, etc.), good enough to get the point across, and boring. You know, the stuff we're grown up being taught how to write for essays by our English teachers.

AIs are trained to write using common writing, from boring recipe blog posts to so-so prose on fan fiction sites. There are some good writers in there--it's why you can say "make a story in the style of X"--but the vast majority of what it works off of is mediocre, at best. And given that's it's a probabilistic engine trying to decide what word should go next based on what came before, it's no surprise that that's what it churns out.

 

[MillieDynamite]

Because it won't be published if doesn't pass. Perhaps the answer is they find a different site to publish on.

I do not think it would behoove an author to do that because nth showed that the detectors suck. Why should anyone have to change their writing to pass a crappy inaccurate detector? nth showed where one of the most popular stories here got flagged by a detector for being one hundred percent ai and the story has been up on the site for almost 20 years lol. I think what we learned is that nobody should be using them and also we have no idea if the site owner is using a detector or their own judgement. They told one person their story "still reads like ai" so maybe we should not be telling people what to do with their stories when we don't even know how or why they are getting rejected. Running peoples stuff through these detectors just adds to the witch hunt mentality

 

[alohadave]

Bonus fun: most of the companies that provide AI detectors also make tools to help evade AI detectors.

Sell the problem and sell the solution to the problem.

 

[MillieDynamite]

It uses over complicated sentances, which is more of a sin than the flower prose. It reads like a romance writer from the 18th or 19th century. So, for shits and Giggles, I asked it to write in the Style of Raymond Chandler, this is better.

rewriter the above in they style of Raymond Chandler
ChatGPT
Meet Emily, a dame with a grin as wide as the city skyline at sunset. She's the kind of gal who could turn a gray day into a Technicolor dream, a walking burst of sunshine in a noir world. With laughter that echoes like jazz notes in a smoky joint, Emily's got a way of making the shadows scatter and the blues take a backseat.

This dame's positivity isn't just a facade – it's a part of her DNA. Her smile ain't just a mask; it's a beacon that can cut through the darkest alleys. Emily's got this magnetic pull, a charisma that draws folks in like moths to a flame. Whether she's spinning tales of her latest caper or just shootin' the breeze, Emily's got that rare knack of sprinkling a little joy dust wherever she goes. In a city full of hard knocks and tough breaks, Emily's a living reminder that happiness ain't just a luxury – it's a necessity.

With a little work, it would be useable. Notice I misspelled the as they and still figured out what meant. It shows fewer errors.

Even so, it still doesn't feel right.

See, I don't consider this dreadful. I just don't consider it good. I could EASILY see this as part of someone's first story; the writing in After the Future is Gone is only marginally better, if I'm being 100% honest. I think that's the problem: AI detectors don't detect bad writing; they detect common writing: technically competent (in terms of grammar, spelling, etc.), good enough to get the point across, and boring. You know, the stuff we're grown up being taught how to write for essays by our English teachers.

AIs are trained to write using common writing, from boring recipe blog posts to so-so prose on fan fiction sites. There are some good writers in there--it's why you can say "make a story in the style of X"--but the vast majority of what it works off of is mediocre, at best. And given that's it's a probabilistic engine trying to decide what word should go next based on what came before, it's no surprise that that's what it churns out.

 

[MillieDynamite]

Well, you're more than capable of showing us how you would have it write a limited introduction.

Your prompt is lousy