The AI Rejection Conversation Matters

[Britva415]

I don't know about that. Most of the complaints seem to be about AI writing a flat, monotonous, dead style. See my post above, about every sentence having the same structure without any effort to flow from one to another.

That might be something which Laurel may use to sense whether she thinks it’s AI-written or not, but my point was that the style isn’t what she objects to, it’s the technology.

If she’s rejecting anything for “style” or for a deficit of style, you’d never know it from some of what gets approved every day.

 

[SaraLab]

Initial AI text detection seems to occur within a few hours to a few days of a story being submitted. So I wonder if AI rejects at 1-3 weeks after submission are the result Laurel manually flagging them based on what she thinks.

In which case, that would mean that I have to somehow rewrite them to her liking with zero feedback.

 

[SaraLab]

I've reached out to couple editors to see if they would be willing to alter my writing and/or add in small parts to help break up whatever patterns are triggering the detectors or Laurel. At this point, I think that might be my best option if I want to have these stories published. I care more about the content than my writing style.

 

[DrPope]

I'm curious how people have responded to the rejection, what they've done, and how their efforts have been responded to

Whilst I have never been rejected for AI (never submitted anything for over a decade) here I have also never had a text flagged as AI on any tool I’ve used to see how it worked or anything I’ve published to Amazon. I have seen several really good authors (including one of the best in years) leave this site despondent after numerous false AI accusations by whatever bullshit tools the monkeys that run that disgusting undocumented program use. I call for transparency and communication from the sites admins…. To deaf ears. In the meantime all I can do is remove Literotica from advertising placement for the very few options I have open to my control. It’s not much … maybe only a few hundred bucks but it’s a few hundred bucks these despicable people won’t be getting.

 

[Plathfan]

Out of curiosity, I input a 15-sentence passage into Pharly.AI and received a 14% likelihood of it being written by AI. These are the three sentences that were flagged:

A sharp scent of antiseptic lingered in the air.

He lay there, eyes half-closed, cocooned in white bandages that covered his entire body, leaving only his eyes, mouth, and nostrils exposed. Wires and tubes connecting him to monitors emitted steady, rhythmic beeps.

Why on earth, I wondered, these specific sentences were flagged and not the others? What’s AI-like about “A sharp scent of antiseptic lingered in the air”?

But then I changed lingered to hung and removed tubes, and voila, 100% human! Just two words... and these ridiculously stupid machines are deciding our fates.

 

[MillieDynamite]

Lingered in the air would trigger a flag. The complexity of the two sentences would also be a flag if they fell close to Lingered in the air. Not so much if there is a lot of text between the Lingered and them. Anything rhythmic can send up a flag.

Out of curiosity, I input a 15-sentence passage into Pharly.AI and received a 14% likelihood of it being written by AI. These are the three sentences that were flagged:

A sharp scent of antiseptic lingered in the air.

He lay there, eyes half-closed, cocooned in white bandages that covered his entire body, leaving only his eyes, mouth, and nostrils exposed. Wires and tubes connecting him to monitors emitted steady, rhythmic beeps.

Why on earth, I wondered, these specific sentences were flagged and not the others? What’s AI-like about “A sharp scent of antiseptic lingered in the air”?

But then I changed lingered to hung and removed tubes, and voila, 100% human! Just two words... and these ridiculously stupid machines are deciding our fates.

 

[Thanatos_X13]

Out of curiosity, I input a 15-sentence passage into Pharly.AI and received a 14% likelihood of it being written by AI. These are the three sentences that were flagged:

A sharp scent of antiseptic lingered in the air.

He lay there, eyes half-closed, cocooned in white bandages that covered his entire body, leaving only his eyes, mouth, and nostrils exposed. Wires and tubes connecting him to monitors emitted steady, rhythmic beeps.

Why on earth, I wondered, these specific sentences were flagged and not the others? What’s AI-like about “A sharp scent of antiseptic lingered in the air”?

But then I changed lingered to hung and removed tubes, and voila, 100% human! Just two words... and these ridiculously stupid machines are deciding our fates.

Yup, and to make matters worse, you can "fix" the flagged sentences, and then have other sentences that weren't flagged before suddenly get flagged when you run it again. So if you're trying to combat this "issue," you end up playing the world's most ridiculous and tedious game of whack-a-mole. It's profoundly stupid. --_--

 

[StillStunned]

Some of these AI detectors remind me of Baltar in "Battlestar Galactica", just randomly identifying people as Cylons and hoping he gets lucky.

 

[SaraLab]

Just got another AI rejection, and its still not clear what I need to change to avoid that.

 

[Britva415]

"tubes"

 

[onehitwanda]

Some of these AI detectors remind me of Baltar in "Battlestar Galactica", just randomly identifying people as Cylons and hoping he gets lucky.

I am going to develop a AI detector based on the random results of rolling 6 d6 and choosing 5 at random. Two identical numbers and you're an AI, boyo.

I'll make millions.

 

[AwkwardlySet]

I am going to develop a AI detector based on the random results of rolling 6 d6

Reading this, the first thought that came to my mind was "What feeble weapon rolls a d6?... Shortsword? Handaxe? Sickle? Club?" There should be a pill that treats severe cases of nerdism.

 

[StillStunned]

Reading this, the first thought that came to my mind was "What feeble weapon rolls a d6?... Shortsword? Handaxe? Sickle? Club?" There should be a pill that treats severe cases of nerdism.

Fireball?

 

[onehitwanda]

Reading this, the first thought that came to my mind was "What feeble weapon rolls a d6?... Shortsword? Handaxe? Sickle? Club?" There should be a pill that treats severe cases of nerdism.

mace and flail both roll 1d6 in 2nd edition :/ morningstar was 1d8 I think.

 

[onehitwanda]

Fireball?

1d6 per level, max 10d6, save vs spell for half damage.

 

[StillStunned]

1d6 per level, max 10d6, save vs spell for half damage.

So, we've learned that Wanda is a level 6 wizard or sorcerer.

 

[AwkwardlySet]

1d6 per level, max 10d6, save vs spell for half damage.

Then there is the burning effect that does damage per turn... double damage if the body is oiled.

 

[onehitwanda]

So, we've learned that Wanda is a level 6 wizard or sorcerer.

I always played paladins. mainly because they got to lay hands on everyone...

 

[HeyAll]

This may be helpful to many:

https://twitter.com/x/status/1864550214562140165

It's a long post showing what AI writing tends to do, how it formulates sentences and what words they use.

 

[AG31]

This may be helpful to many:

https://twitter.com/x/status/1864550214562140165

It's a long post showing what AI writing tends to do, how it formulates sentences and what words they use.

Thanks very much!

 

[secretLace]

As one who has been rejected as well, that was a great read HeyAll. Without communication from Literotica it is what it is. Sad, but Laurel must deem it nessecary otherwise it wouldn't be used. Most places in higher education have stopped using AI to find AI since it has been proven that it does not work.

Like most who have been rejected I would love feedback but if you haven't used AI to write then I would be 'changing your writing' to fit a mould. I hope that this blows over soon so that we can go back to publish smut like we have always done.

 

[TPS23]

My latest submission got flagged for suspected AI use. I've re-submitted with an expanded admin note, but based on the experiences of others, I'm expecting to be asked to make edits in the dark to attempt to make it past the AI gatekeeper.

Even more unpleasantly, the rejection seemed to trigger a chain reaction. My three other stories (previously published with no issues) have been retrospectively rejected for the same reason. It all happened within minutes of my clicking publish, so surely automated. Has anyone seen that before? I've messaged Laurel, but again, I'm not hopeful of a reply based on what I've read on here.

The return of my back catalogue worries me. I've had feedback on those already, so I have no drive to re-edit them to meet Lit's AI's satisfaction. I wonder though, if I make edits and have my new story accepted, will it get rejected again at the time of my next submission (or when the AI detection programme gets an update)?

 

[StillStunned]

I'm sorry to hear it, especially about existing stories being taken down! I can't offer any advice, though, only my sympathy.

 

[AG31]

My latest submission got flagged for suspected AI use. I've re-submitted with an expanded admin note, but based on the experiences of others, I'm expecting to be asked to make edits in the dark to attempt to make it past the AI gatekeeper.

Even more unpleasantly, the rejection seemed to trigger a chain reaction. My three other stories (previously published with no issues) have been retrospectively rejected for the same reason. It all happened within minutes of my clicking publish, so surely automated. Has anyone seen that before? I've messaged Laurel, but again, I'm not hopeful of a reply based on what I've read on here.

The return of my back catalogue worries me. I've had feedback on those already, so I have no drive to re-edit them to meet Lit's AI's satisfaction. I wonder though, if I make edits and have my new story accepted, will it get rejected again at the time of my next submission (or when the AI detection programme gets an update)?

Wo! That's doubly scary. But now that I've read some reports about various publishing entities being flooded with 100s of submissions more than normal, I do see that it's a problem that needs to be fought. But as AI gets better and better, I don't see what the strategy should be.

 

[MillieDynamite]

They pull stories that have been in publication from complaints of the reader suspecting AI and only pull them after they "REVIEW" the story; however, it is that they do their check.