"Sent Back"

[onehitwanda]

seems Grammarly is feeling the heat: https://www.grammarly.com/authorship

 

[EmilyMiller]

seems Grammarly is feeling the heat: https://www.grammarly.com/authorship

I don’t always take any notice of what humans suggest to me, let alone mindless statistical inference engines…

 

[onehitwanda]

I don’t always take any notice of what humans suggest to me, let alone mindless statistical inference engines…

Soon you will join me in preemptively hating all humans because it saves time...

Also, the ULA has decided that this thread needs to be derailed.

 

[EmilyMiller]

Soon you will join me in preemptively hating all humans because it saves time...

Also, the ULA has decided that this thread needs to be derailed.

View attachment 2630240

And the prize for GIF most used by a famous writer of erotica goes to…

I could keep pulling on the strings until it created a black hole without any other discernible result

 

[lasoldier]

I also got hit with the blank bullet point rejection. An image so you can all see it for yourself:

I guess I'll send a note...but man, I'm tired.

 

[MartaMaine]

It won’t be for spelling. Grammarly has now incorporated an LLM (aka genAI). Using it for spelling and grammar is fine, many authors do. But accepting any suggestions it makes as to how to reword your sentences leaves you wide open to the risk of AI-based rejection as it’s using AI to make those suggestions. Whatever Lit uses seems very sensitive to Grammarly-generated text.

Thanks Emily. I've resubmitted my story with a note stating that Grammarly has only been used for spelling and grammar checks. Suggestions for rewording are ignored.

Strange thing is that I've been using Grammarly in this way for years. In fact, I uploaded a story to Lit a couple of weeks ago without issue.

 

[EmilyMiller]

Thanks Emily. I've resubmitted my story with a note stating that Grammarly has only been used for spelling and grammar checks. Suggestions for rewording are ignored.

Strange thing is that I've been using Grammarly in this way for years. In fact, I uploaded a story to Lit a couple of weeks ago without issue.

There are also random glitches in most AI detectors, I hope that’s the case with you. Good luck

 

[MartaMaine]

Resubmitted Tuesday, still in Pending!

 

[EmilyMiller]

Resubmitted Tuesday, still in Pending!

Yeah - that feels problematic again. Sorry to hear .

 

[Britva415]

There's some chatter going around that I'll repeat, to the effect of:

Using an automated tool like Grammarly to find spelling and grammar mistakes is thought to potentially have different outcomes, AI-detection-wise, depending on whether you allow the tool to automatically make the correction for you or you simply type the correction yourself after the tool shows it to you.

 

[vootvoot]

Most of the free AI checkers are super unreliable. Pangram’s rep is that it’s good at flagging actual AI, but can sometimes give false positives. I.e. if it’s AI-written, it will be quite good at flagging that. But if it’s human-written, it’s not so good at flagging that.

The result is you weed out virtually all AI-written stuff, at the expense of falsely flagging some innocent stories.

—

DISCLOSURE: I’ve run excerpts from quite a few of my published stories and the same for my WIPs through Pangram and got 100% human-written with high confidence every time. If Pangram says it’s human, it’s very likely human. If it says it’s AI, it could be AI, or it could be making a mistake.

My understanding is that it's the exact opposite. They try very hard to avoid falsely accusing human writing of being AI generated, at the expense of sometimes failing to correctly flag some AI generated content. See their blog post on the subject:

In AI detection, a false positive is far worse than a false negative. Repeatedly accusing students who write assignments by themselves with no AI assistance of AI plagiarism greatly undermines trust between student and teacher, and can cause a great deal of anxiety and stress for the student. On the other hand, a false negative may mean that a cheater may slip through every once in a while, which is not as bad of an outcome for an AI detection tool.

https://www.pangram.com/blog/all-about-false-positives-in-ai-detectors

 

[EmilyMiller]

My understanding is that it's the exact opposite. They try very hard to avoid falsely accusing human writing of being AI generated, at the expense of sometimes failing to correctly flag some AI generated content. See their blog post on the subject:

https://www.pangram.com/blog/all-about-false-positives-in-ai-detectors

I was going on using the thing, not their marketing material…

 

[RejectReality]

There's no such thing as an AI detector that's any better than tossing a coin. They have to show results to sell their product, and that means finding AI, which means it's going to be the default to find AI, no matter what their propaganda says.

 

[vootvoot]

I was going on using the thing, not their marketing material…

From your previous comments, I was under the impression you'd tried it with various samples of your own writing and never had it flagged. Are you saying you've found examples of human writing that are falsely flagged as AI by Pangram? I'd be curious to see them!

 

[MartaMaine]

It's all very confusing. I resubmitted my story 4 days ago, along with 2 others. The original one is still in pending. Of the other 2, 1 has been published, and the other has been sent back with the same message regarding the use of Grammarly.

 

[millamomud]

Most of the free AI checkers are super unreliable. Pangram’s rep is that it’s good at flagging actual AI, but can sometimes give false positives. I.e. if it’s AI-written, it will be quite good at flagging that. But if it’s human-written, it’s not so good at flagging that.

The result is you weed out virtually all AI-written stuff, at the expense of falsely flagging some innocent stories.

—

DISCLOSURE: I’ve run excerpts from quite a few of my published stories and the same for my WIPs through Pangram and got 100% human-written with high confidence every time. If Pangram says it’s human, it’s very likely human. If it says it’s AI, it could be AI, or it could be making a mistake.

This is from my current rejection on here, so... ¯\_(ツ)_/¯

 

[MartaMaine]

View attachment 2634556


This is from my current rejection on here, so... ¯\_(ツ)_/¯

Yep, same here. I used the free trial of Pangram to check all my stories, and they all came back as 100% Human written, yet I have now had 2 of them rejected.

 

[SmilingLez]

Yep, same here. I used the free trial of Pangram to check all my stories, and they all came back as 100% Human written, yet I have now had 2 of them rejected.

Sometimes Literotica’s AI filter rejects stories even when outside detectors say the writing is 100% human. That usually means the problem isn’t the author or the writing style - it’s the backend.

https://forum.literotica.com/threads/backend-misclassification-event-bug.1656201/


https://forum.literotica.com/threads/systemic-backend-failure-–-multiple-symptoms.1656323/

Literotica’s tool is tied into an older part of their publishing system, and when that system glitches, it throws false positives. So if your story keeps getting flagged, it’s most likely the site’s internal tech misfiring, not anything you did wrong.



As always my opinion and YMMV

 

[MartaMaine]

Well, I have 2 stories that I resubmitted (including the one that started this thread off), and both have been in pending for over a week. I wonder if they'll ever be approved.

 

[SkyBubble]

If the PM to Laurel doesn't work, you might need to post this in the Editors Forum and see if an editor can help you.

 

[MartaMaine]

Yay! Life Model (the first rejection) has been published!

 

[dmallord]

Got 4 rejection notices today on a four-part story. AI, of course. In the Editor’s notes field, I’d entered a notice that I used Grammarly for spelling and punctuation, not for editing.

So a note doesn’t mean squat, does it?

Edit what exactly to make it seem less AI?

Maybe I should tell Grammarly to rewrite the whole 60K story so it doesn’t sound AI and resubmit it, saying I had the AI create a non-AI imitation of me? Maybe circular logic will let that fly?

I’m done–no energy left to pursue this.

~~~~~

So, I opened Grammarly and asked, “How’s this set with you?” Pasted the above and this is it’s reply:

It reads as raw, frustrated, and exhausted—which makes the emotion feel very real. If that’s the point, it works.

If you want it to land more clearly and less like a spiral of sarcasm, I’d soften a few lines and tighten the message. For example:

I received four rejection notices today for a four-part story, presumably because of AI concerns. In the notes field, I had disclosed that I used Grammarly for spelling and punctuation only, not for editing.

So what, exactly, are they expecting me to change?

The idea of asking Grammarly to rewrite the entire 60K story so it sounds less AI-generated feels absurd—almost circular logic.

At this point, I’m done. I don’t have the energy to keep pursuing it.


If you want, I can also make it sound:

• more polished
• more angry/raw
• more professional and restrained

~~~

NOPE, didn’t want that version, I liked mine just the way it is. So was my story’s version, fixed for spelling and punctionatin of course cause as my age I cn’t hardly remember how to spell ‘spit.'

NOW, I’m done.

 

[MartaMaine]

Got 4 rejection notices today on a four-part story. AI, of course. In the Editor’s notes field, I’d entered a notice that I used Grammarly for spelling and punctuation, not for editing.

So a note doesn’t mean squat, does it?

Edit what exactly to make it seem less AI?

Maybe I should tell Grammarly to rewrite the whole 60K story so it doesn’t sound AI and resubmit it, saying I had the AI create a non-AI imitation of me? Maybe circular logic will let that fly?

I’m done–no energy left to pursue this.

~~~~~

So, I opened Grammarly and asked, “How’s this set with you?” Pasted the above and this is it’s reply:

It reads as raw, frustrated, and exhausted—which makes the emotion feel very real. If that’s the point, it works.

If you want it to land more clearly and less like a spiral of sarcasm, I’d soften a few lines and tighten the message. For example:

I received four rejection notices today for a four-part story, presumably because of AI concerns. In the notes field, I had disclosed that I used Grammarly for spelling and punctuation only, not for editing.

So what, exactly, are they expecting me to change?

The idea of asking Grammarly to rewrite the entire 60K story so it sounds less AI-generated feels absurd—almost circular logic.

At this point, I’m done. I don’t have the energy to keep pursuing it.


If you want, I can also make it sound:

• more polished
• more angry/raw
• more professional and restrained

~~~

NOPE, didn’t want that version, I liked mine just the way it is. So was my story’s version, fixed for spelling and punctionatin of course cause as my age I cn’t hardly remember how to spell ‘spit.'

NOW, I’m done.

Keep plugging away. All my stories are now published.