"AI" Rejection

[MillieDynamite]

I just this minute, used this site to check my open to my latest story.

 

[dmallord]

Not that we should have to change our writing styles, but some of the detection services highlight exactly what they think looks AI generated. This one for example: https://contentatscale.ai/ai-content-detector/

I tried this using a story I am currently writing. It limited me to some 400 words, but it said it was probably human-written. Glad to know - I'm human or at least some AI thinks I am.

 

[MillieDynamite]

If the highlights are green, it's fine; yellow is not that bad; orange means trouble; and red means you're dead.

That one gave me a completely different answer ... Had a few lines here and there, but not whole passages which seems to be what the rejection is about and what the other AI detector told me. Just goes to show how false these things can be if they all give different answers

 

[XerXesXu]

You should take it as it ain't getting posted until you get that green. Looks to be written by a Human at the top.

AI Detector This is the one my editor uses sometimes, not on mine, but on a person who uses AI.

An AI program that’ll rewrite text, not written by an AI program, creating text written by an AI program that appears to be text written by a human – for a very reasonable fee.

If a rival comes up with a detector that detects text written by all the usual suspects, and AI Detector, will they refund that reasonable fee?

 

[SyleusSnow]

An AI program that’ll rewrite text, not written by an AI program, creating text written by an AI program that appears to be text written by a human – for a very reasonable fee.

If a rival comes up with a detector that detects text written by all the usual suspects, and AI Detector, will they refund that reasonable fee?

Yep, https://contentatscale.ai/ai-content-detector/ offers a "paraphrase" service to rewrite your AI-writing to be undetectable by AI detectors. For a reasonable fee.

It's turtles bullshit all the way up and all the way down.

And remember: if your AI writing tool wasn't produced by a high-tech firm located in the Silicon Valley region of the US, legally it can't be called AI. It's only Sparkling Bullshit

 

[ProperFatherlyLove]

https://arstechnica.com/information...i-admits-that-ai-writing-detectors-dont-work/

 

[MillieDynamite]

Well, no, I wouldn't use it that way. If the AI says it's AI (and it's big enough to matter to me) I rewrite it. I don't always check my work. You have to do 2500 characters at a time, and that takes a long time even on 3500 words.

An AI program that’ll rewrite text, not written by an AI program, creating text written by an AI program that appears to be text written by a human – for a very reasonable fee.

If a rival comes up with a detector that detects text written by all the usual suspects, and AI Detector, will they refund that reasonable fee?

 

[djrip]

Well, no, I wouldn't use it that way. If the AI says it's AI (and it's big enough to matter to me) I rewrite it. I don't always check my work. You have to do 2500 characters at a time, and that takes a long time even on 3500 words.

So what you're saying is, you change the human written words that you crafted, to something else, until an AI says it approves? And that's supposed to be more human and less AI?

 

[lovecraft68]

I don't know much about AI, but if this detector is searching the net to see if this has been written before than hasn't everything been written at some point?

For example, if I take an e-book I published last year and decide to upload it here, is it going to flag because a detector has found it all on the net because its been online before?

Or is there something else it looks for?

In general AI is a nightmare for a variety of reasons, and when it comes to writing and art, its usage is a disgrace and I can't believe people who use it call themselves creators.

 

[djrip]

I don't know much about AI, but if this detector is searching the net to see if this has been written before than hasn't everything been written at some point?

For example, if I take an e-book I published last year and decide to upload it here, is it going to flag because a detector has found it all on the net because its been online before?

Or is there something else it looks for?

In general AI is a nightmare for a variety of reasons, and when it comes to writing and art, its usage is a disgrace and I can't believe people who use it call themselves creators.

I don't know for sure but I don't think it's searching for matches anywhere; I think it's using AI to decide if the text sounds like something AI would come up with. Very badly.

 

[MillieDynamite]

Only if I feel a need to, normally, I don't fool with it. I mean, what is one fucking line, teen to fourteen words, out of 6,000? If I think it doesn't read well, I change it. If I don't think it needs changing, fuck that shit. But if every line is highlighted as AI you have a problem with your writing.

So what you're saying is, you change the human written words that you crafted, to something else, until an AI says it approves? And that's supposed to be more human and less AI?

 

[lovecraft68]

I don't know for sure but I don't think it's searching for matches anywhere; I think it's using AI to decide if the text sounds like something AI would come up with. Very badly.

No wonder this is a mess.
I think a human editor could pick up on AI, but if laurel is using a program of some sort, then this is just going to get worse.

 

[djrip]

Only if I feel a need to, normally, I don't fool with it. I mean, what is one fucking line, teen to fourteen words, out of 6,000? If I think it doesn't read well, I change it. If I don't think it does, fuck that shit. But if every line is highlighted as AI you have a problem with your writing.

I finally tried running one of mine through a detector, and got less than 1% likely to be AI. Maybe I am a real boy after all...

 

[MillieDynamite]

Is your nose getting longer, @djrip? Yes, it is. Hum, maybe you're not a real boy!

I finally tried running one of mine through a detector, and got less than 1% likely to be AI. Maybe I am a real boy after all...

 

[SoCalAshley]

No wonder this is a mess.
I think a human editor could pick up on AI, but if laurel is using a program of some sort, then this is just going to get worse.

I don’t know for a fact that Laurel is using an AI detector because she has not responded to this forum or to my message. I do know that my latest story, as well as many other authors' stories, has been rejected, so yes, it's a huge mess and it is going to get worse.

 

[lovecraft68]

I don’t know for a fact that Laurel is using an AI detector because she has not responded to this forum or to my message.

And this is the problem no one here has the stones to mention. This is a big deal, she knows it is, she knows stories are being rejected unfairly, knows there's been discussion of it here but....

Of course its to much to ask that she comes down from on high and addresses it. I mean the authors here are the reason the site makes whatever money it does, why would she post even so much as a "hey, we're working on it"

Manu is no better, he shows up looking for feedback on whatever upgrade no one asked for but is nowhere to be found when there's an issue.

 

[theWollstonecraftWoman]

On my most recent rejection, the message included this: 'This chapter appears to lean heavily on generative AI.'

It's the 'appears' that stands out to me. I've said 3 times now I don't use AI and yet they keep rejecting it because it 'appears' to have AI content. This was a part that took almost a month to be uploaded the first time - why would it take me a month to use AI to write a chapter?

This fourth rejection was actually the worst one because of that 'appears' Laughing through the pain

How on earth do you do the lines though those words? I can't find that anywhere. I feel so stupid sometimes.

 

[Backagainpart4]

How on earth do you do the lines though those words? I can't find that anywhere. I feel so stupid sometimes.

While the text is highlighted, on the editor toolbar click on the three dots next to the color palate. Next to the Font chooser is an S with a line through it. Click on that to put a line through the text.

 

[alohadave]

How on earth do you do the lines though those words? I can't find that anywhere. I feel so stupid sometimes.

 

[MillieDynamite]

Well, hells bells, I knew, didn't know that.
@alohadave

 

[DaedalusX]

It seems I'm not the only one facing this bizarre situation. About a week ago, I submitted four stories, only to have the first one rejected earlier today on the grounds of "AI use." Intrigued by this, I decided to conduct a little experiment.

I put each of my stories through what are considered the top five free AI detectors to see how they would fare. The outcomes were very inconsistent.

Out of the five detectors, three identified my human-written stories correctly, while two mistook them for AI-generated content. Curiously, the random nonsensical text I created using ChatGPT received a similar split verdict – three detectors labeled it as human, while two flagged it as AI. It appears to be essentially random.

So, I moved forward and resubmitted all four stories, this time explicitly stating in each that no AI was involved in their creation (since that was the most common advice given by others). If they're rejected again, so be it. My purpose here is to engage in the joy of sharing and reading stories, not to endlessly navigate bureaucratic hurdles. Fuck that.

I'm half-tempted to conduct a 'shits and giggles' experiment: take photos of my handwritten stories, (written on real dried wood pulp stuff!) record the entire process of typing them into Word, and then submit them. I have a sneaking suspicion that even this would still result in rejection due to supposed AI use. It would be a fun test.

Regarding the few discussions about writing ability being a potential trigger for these rejections, I believe that aspect is completely irrelevant. Whether your story has a Flesch Reading Ease score of 5 or 95, it doesn't matter. The essence of this platform is to be a haven for creative writing and the sharing of stories on the interwebs, specifically - smut. Don't get it twisted.

 

[Andreas_Kreuz]

Here on a small scale you have an example why the fear of AI is justified. AI is allowed to make decisions without human supervision. And its faults mess up real peoples lives.

 

[Jackie.Hikaru]

I wrote an entire series here about humans being at the mercy of a high-ranking robot. Now that one of my stories has been rejected with the false claim that it was written by AI, I find myself at the mercy of a high-ranking AI detection bot. Definite glitch in the simulation.

When reality reflects art.

 

[u1u1u1]

My conversations with Laurel (my author's notes and her rejection notes) strongly suggest that she is not rejecting things based solely - or even primarily - on the output of an AI detection bot. If she's routinely using a bot, then she's also checking to see whether they "read as Ai" for herself.

For what it's worth, the back and forth on my latest story has now been:
Me (initial submission): I have not used AI.
Laurel: please say what programme you used to generate this.
Me: I wrote it in a word processor, checked it with three AI detection bots, and then ran it through LanguageTool to double-check the grammar and spelling.
Laurel: Are you sure you didn't use tools like Grammarly? Parts of it read like AI to me.
Me: I didn't use Grammarly. As I said, I use LanguageTool.
Laurel: "If you are using a grammar check program sparingly (as a spellcheck, to fix punctuation, and/or as a thesaurus), that is fine. If you are allowing a grammar check program to “rewrite” your words, then you are using AI." <- this is an exact quote, all the rest is paraphrasing.
Me: I haven't, I wrote this all myself.

 

[djrip]

My conversations with Laurel (my author's notes and her rejection notes) strongly suggest that she is not rejecting things based solely - or even primarily - on the output of an AI detection bot. If she's routinely using a bot, then she's also checking to see whether they "read as Ai" for herself.

For what it's worth, the back and forth on my latest story has now been:
Me (initial submission): I have not used AI.
Laurel: please say what programme you used to generate this.
Me: I wrote it in a word processor, checked it with three AI detection bots, and then ran it through LanguageTool to double-check the grammar and spelling.
Laurel: Are you sure you didn't use tools like Grammarly? Parts of it read like AI to me.
Me: I didn't use Grammarly. As I said, I use LanguageTool.
Laurel: "If you are using a grammar check program sparingly (as a spellcheck, to fix punctuation, and/or as a thesaurus), that is fine. If you are allowing a grammar check program to “rewrite” your words, then you are using AI." <- this is an exact quote, all the rest is paraphrasing.
Me: I haven't, I wrote this all myself.

I can't stand tools suggesting alternative phrasings as I type. Because then I sit there thinking about do I want that change? Or do I want to keep what I had? Maybe I like it that way, but now that I've seen it suggested I almost don't want it, just because I don't want to change it because of the suggestion. Or I don't want it, but every time I scroll past that underline catches my eye and I have to check it and decide to ignore it again. So I end up arguing with my word processor and second guessing myself instead of thinking about what I want to write.