AI Allegations Thread

[DDX84]

lolol you're the one who brought up the contest saying people should be mad nth is using ai tools for them.

No, as you'll see below.

On that note, how do you all feel about entering into a contest here on Literotica with someone who's just freely admitted to doing exactly that, while creating stories on an industrial basis?

I asked the wider forum how they felt about the prospect of entering into a Literotica contest with someone who freely admits to the use of software tools and AI to contribute towards their final drafts. Users can choose to answer if they wish. Some may be perfectly fine with that, others not.

I'm very much against it and my view is that anyone in NoTalentHack's position should be banned from future contests here.

 

[NoTalentHack]

A professional editor won't be changing words all on his/her own. They'll question word usage and maybe even suggest better wording, but it will be left to the author to make final choices--retaining responsibility and author's voice.

And that's exactly what I do. I don't talk about the software I use in the forewords or afterwords of my stories--although some people in LW do mention if they use Grammarly or whatever--because it just doesn't matter. I credit my beta readers, though, because they're people and I want them to know how much I value them and their friendship. Maybe I should mention PWA, but most people aren't going to give a shit, so who cares? Only absolute tools like this guy would, unless they want to know about the process, and those folks are welcome to message me. Obviously, I don't think there's anything wrong with what I'm doing, so...

 

[DDX84]

And that's exactly what I do. I don't talk about the software I use in the forewords or afterwords of my stories

For whatever reason, you decided to talk about them here. You made the disclosures on a public forum filled with your peers.

Are those who are opposed to your use of software tools supposed to turn a blind eye because you want to take the disclosures back?

My views on what you've done are clear. If Literotica has no problem with your use of software to contribute to your stories, then I consider the issue closed. I'd then expect Literotica, at some point, to amend their terms and conditions to make it abundantly clear what they consider acceptable, and what they don't.

As to why the moderators are permitting you to swear at me, twice now, is up to them. Just don't think that I consider you as anything other than a cheat.

 

[XerXesXu]

No, the latter doesn't necessarily follow from the former. The other possibility is that the rewrite constitutes a derivative work, in which both the original author and the rewriter have copyright interest, and neither would be able to post it to Literotica without the other's permission.




Hmmmm.



There are very minor differences in wording from the version I have available, but nothing beyond what one might expect from translation choices. It definitely has "learned" the text very closely, if not verbatim, though the quote is shorter than the whole chapter I asked for.



That one is verbatim to the copy I have at hand.



whoops!

Not that an excerpt that short would be likely to constitute copyvio, but hopefully that illustrates that GPT can indeed learn the source material so closely as to be effectively "making a copy". I expect most of the stories it's trained upon would be less perfectly memorised, but it doesn't need to be perfect to be copyright violation.



"Essentially" and "fancy version" are doing a lot of heavy lifting here. 8letters' statistical analyses don't have the capability to reproduce entire chunks of the stories they're based on.



Which, as seen above, they can do.

Please note that ChatGPT is a Search Engine bot - an LLM that can act as a search engine. As an LLM it retains no copy of the training set. The content of the training set can be found by searching the internet because it was trained on the internet. Be sure to distinguish between each of these characteristics. The bot says that 'The Hobbit' is in the public domain, it is in some jurisdictions. Is it still in copyright in some jurisdictions? Yes, but you have to specify your prompt appropriately. As an LLM, does it retain a 'memory' of training? No. That doesn't mean it can't find a copy of 'The Hobbit' on the internet, as you would if you didn't bother with these new-fangled productivity tools and did your internet search the old-fashioned way.

 

[DDX84]

Again, reading comprehension, you dumb fuck.

Until then, fuck right off,

A real upstanding member of the community here, folks. I've lurked here for a few years, why is it always you that starts swearing at people in these threads? How many times is that now in the last year?

What value are you providing to this community of authors behaving like this?

I'm glad though that OverconfidentSarcasm chose to highlight you. You're proof to everyone who wants to publish on Literotica that you can quickly acquire 2,000 followers from the depths of Loving Wives, while offering nothing to them but hastily-produced cut-and-shuts, desperately pieced together by your raft of software and the meagre contents of your own head.

I like my stupid, lazy phrasing.

At least you're not pretending to be a competent writer.

I read so much of your stuff last week, it's not good at all. Dropped from the peloton. Just like MillieDynamite said a few posts ago, when she pointed out how bad your work was, not knowing it was you (ouch, that has to sting). And that's with you relying on the software to do your absolute best. God only knows just what standard you'd produce if you were sat in isolation, with just a Microsoft Word document to rely upon. The one thing I can say with 100% certainty is that you couldn't do it.

Talks about his full-time job as a software developer, but has the time and energy to write over 1.5m words in one year, after all those long shifts, across 51 different stories, with screenshots of him accessing a generative-fill panel that he "totally didn't get tempted to use, trust me guys...". But just spoke about how he doesn't have the energy to even edit a document, hence his use of the software and AI in the first place. Just how naive do you think the people are here? The reality here is plain as day.

The other authors must've already sensed the truth over the last year of you being here. You come to these forums and you act like a thug. Didn't you once invite @lovecraft68 to fight you somewhere in Texas? That was in the summer, right? Because he dared to talk down to your Loving Wives knuckle-draggers? What were you thinking? Actually, I have to find that quote, it's too good. I'll find it, I will.

I swear, that is the most "internet tough guy" statement ever. "I'll fight you, bro!"

Anyways, down in Texas, so no worries.

I better get out of this thread before you threaten to attack me next! I'm a writer, not a fighter, and that's something that you'll never be able to say.

Oh, man, that thread! That's the crazy thread where you're in there bragging about getting that reviewer banned from the website. The Tilan-alt fiasco! It was so crazy that they locked it, I remember that. Here's a great line I've just found, you'll love this.

I have independent reasons for believing they were the same person, but I don’t want to share details. I would put the odds at about 90% that they were

Would you look at that? Our very own NoTalentHack, making a claim that the Stacnash reviewer and Tilan were the same person based on...the balance of probabilities! The irony is actually beautiful. Were they the same person in the end? Does anyone know? Weren't some people pretty annoyed that they got banned just because NoTalentHack lied and made the whole thing up? Some people hated the reviewer too and lots of people hated Tilan.

I'll happily say, "yes, I'm the one who reported him and his alts, and I'm glad I did it."

Happy to admit filing the reports!

Well, by your own standard, you won't have any issue with me filing a report against you to the main Literotica website, on the balance of probabilities, that you cheated by using AI to help write your stories? I mean, just like you, I'm about 90% sure.**

Alt hunters do more damage to online communities than any troll ever could.

When it comes to you, NoTalentHack, this quote I found at the beginning of that amazing Stacnash open letter thread nails it. I'm reading it now, you lied and other people paid the price for it. Well, I'm no alt hunter, but I've got no problem reporting a cheat, especially one that's lied since they got here about everything. How do you even keep all those lies straight in your own head? Energy sapping, it's got to be.

Once I've sent my report, I'll send the Stacnash reviewer a feedback email and let them know that NoTalentHack outed himself as a cheat on Literotica. Maybe they'll come back to see it for themselves!

If @Entropy1001001 can just make a new account last Sunday and start posting abusive content here, and NoTalentCheatHack can shout and swear at anyone he wants, it's pretty clear that this place is completely lawless.

** - I'm a lot more than 90% sure that you're a cheat. 90% offered for comedic effect.

 

[TheArsonist]

it's my belief (and it doesn't have to be everyone's!) that ethical, honorable writers of any kind of creative writing conceive of, record, consider, edit, and finalize every word and phrase all by themselves. That's a purist take, I'll readily admit, but it's the only way I've ever found satisfaction while writing.

By that standard, no professionally published author in the 20th century (and few in the centuries before) was ethical and honorable. When you're saying that getting notes on your writing while in draft means it's no longer yours, or that using a thesaurus or spellcheck compromises an author's integrity, that's not just a "purist" take, it's so out-there that I don't think you ought to be involving yourself in the discussion at all: it's like jumping into an argument about sex before marriage or ethical non-monogamy if you believe all sex is immoral—you cannot make a constructive contribution. (You're also calling pretty much everyone else in the thread unethical and dishonorable.)

 

[Bramblethorn]

I just tested the paragraph that got flagged on two other sites, and guess what? It's 100% human-generated content! As it should be.

https://www.scribbr.com/ai-detector/

https://writer.com/ai-content-detector/

Those AI detectors aren't reliable.

As I have already said in this thread. All the more reason why you might not be safe in assuming they're not going to tag you.

 

[Bramblethorn]

A professional editor won't be changing words all on his/her own. They'll question word usage and maybe even suggest better wording, but it will be left to the author to make final choices--retaining responsibility and author's voice.

Yup. Story from my professional editing job:

I was editing a book where the author went out of their way to step on a political button that didn't need to be stepped on. Not going to quote the actual content, but if you imagine you're reading a romance novel with no obvious political component and then one of the characters suddenly out of nowhere says "by the way, China and Taiwan are two separate nations and always will be!" it was about that level of unnecessary looking-for-a-fight.

I flagged this with "hey this is politically controversial and nothing to do with what this book is about, maybe we could just not?"

The author replied, making it very clear that they had Strong Feelings about the topic. So I shrugged, took my money, and let them go on with what I considered to be a very odd and unnecessary decision, because that's the job and it's not my name on the story.

 

[Bramblethorn]

The other authors must've already sensed the truth over the last year of you being here. You come to these forums and you act like a thug. Didn't you once invite @lovecraft68 to fight you somewhere in Texas? That was in the summer, right? Because he dared to talk down to your Loving Wives knuckle-draggers? What were you thinking? Actually, I have to find that quote, it's too good. I'll find it, I will.



I better get out of this thread before you threaten to attack me next! I'm a writer, not a fighter, and that's something that you'll never be able to say.

Erm... that is not a fair representation of what NTH said. Your quote links to the wrong post; here's the actual source.

https://forum.literotica.com/thread...thread-or-is-it.1593945/page-11#post-97453913

LC wrote:

Also, I live in RI, most places in NE I can drive to in a few hours, so if you're even in the area let me know. I'll be happy to have a little meet and greet, and if you want to talk 'tough' that's certainly on the menu, but only if you choose it.

NTH replied:

I swear, that is the most "internet tough guy" statement ever. "I'll fight you, bro!"

Anyways, down in Texas, so no worries. You don't have to puff your chest in person. I can't see spending time in a room with you for any reason, anyways.

Pretty damn obvious he wasn't challenging LC to a fight. The "I'll fight you, bro" was his characterisation of what LC had said. Hence the quote marks.

I disagree with NTH on the ethics of using AI - as I did on his take on Stacnash - but I don't think it's at all productive to take this into such an ad hominem direction and drag in a whole bunch of unrelated stuff.

ETA: I also don't think it's warranted to accuse him of "lying" about Stacnash. As much as I disagree with his assessment that SN and Tilan were the same person, I've seen nothing to indicate that he made that assessment in bad faith. And ultimately it wasn't his decision to ban Stacnash; he's not a mod.

 

[nice90sguy]

No one knows what 'looking like AI' means. How and What AI learns is inaccessible

The first sentence is false.

The second one is broadly true -- it's been a valid criticism of AI as a discovery tool -- and is an extension of an older problem, which I think began in 1974 when brute-force combinatorial algorithm "proved" the hitherto intractable 4-color map theorem.

It's important to remember that the task of generative AI is NOT to "fool people into thinking it was written by a human".

It's to generate good, correct, understanderble, and above all MEANINGFUL text. It does that very well, in its own, PARTICULAR style, which is broadly unremarkable -- as un-jarring and as comfortable to read as possible to most people.

Many of the people who write here (I'm not one of them), also have this same goal in their writing. To make it flow well, and not to let the words "get in the way of" the story. They have a story to tell, and they don't want their readers to focus on the words so much as the story. I'm guessing that it's this type of writer who's going to have problems with false positives from AI detectors. They will tend to do similar things that generative AI does -- not using overly complex sentence structures, use common rather than obscure words, not have extreme variations in sentence length and -- this is a surprisingly important one -- not have any typos in their text! -- Yes, AI detectors regard a lack of typos as evidence of AI assistance.

 

[XerXesXu]

The first sentence is false.

So falsify it. People have opinions on what they think it should be like, gut feelings perhaps. They may be right, but their gut feelings would then be the same as the unconscious and inarticulable knowledge of an AI neural net.

 

[nice90sguy]

Not sure whether you actually read my post, or skimmed it -- there are fingerprints in AI assisted or generative text (admittedly quite blurred fingerprints), some of which I listed in my post.

 

[NoTalentHack]

Assorted nonsense, inflated sense of self-importance, and hurt feelings manifested as one long REEEEEEEEE

Aw, Barney Fife got big mad!

There, there, buddy. It’ll be okay. We’re not going to let the big, bad, mean AI get you. Look, I’ll check your closet and put a nightlight on in your room, okay, champ? Glass of water, too? Okay, sure, sport. Night night now.

 

[Jackie.Hikaru]

the unconscious and inarticulable knowledge of an AI neural net.

alrighty let's pop the hood on this bad boy and see how it makes it's juice...


... Oh.

Am I the artifical intelligence?

 

[TheArsonist]

They will tend to do similar things that generative AI does -- not using overly complex sentence structures, use common rather than obscure words, not have extreme variations in sentence length and -- this is a surprisingly important one -- not have any typos in their text!

Generative AI often produces text with those characteristics, but it's not an inherent property, just part of how the mainstream systems have been tuned. Other systems can have different default characteristics, or prompts can cause them to change their style.

 

[NoTalentHack]

I have a confession to make: I lied about something.

Well, not lied. Forgot, but made a statement in error as a result. I said that I hadn’t had any issues with stories getting through the queue since I started using PWA, but I did. When I submitted Five Nights with Freddy, it came back with an odd note:

• Could the story be developed more fully? That is, is it too short to do the plot justice? Is it under 750 words?

Which was odd, because it was like 8K words, IIRC. I sent a note to Laurel and asked what was up, suggesting that maybe the system hiccuped because of the number of links in it. She responded quickly (I already had an open conversation with her because of the nature of the story and the meta thing seraph and I were doing with alts), saying there was nothing wrong with it, that it might be a glitch in the system, and that she’d send a note to their tech guys. It didn’t even occur to me that it might be related to AI spotting, since the error was so different and, like I said, the whole thing was just a blip that I’d forgotten about.

So maybe I got dinged early on by the AI detector before it got fully integrated? Who knows. That was over a month after I started using PWA, though, and I’d already put three stories through the system in that time, plus all the ones since that have had no issues.

Can’t wait for a certain someone to decide this is additional proof, somehow, that I’m “cheating.” God save us from self-righteous scolds with too much time on their hands and too little understanding of… anything, really.

 

[Voboy]

By that standard, no professionally published author in the 20th century (and few in the centuries before) was ethical and honorable. When you're saying that getting notes on your writing while in draft means it's no longer yours, or that using a thesaurus or spellcheck compromises an author's integrity, that's not just a "purist" take, it's so out-there that I don't think you ought to be involving yourself in the discussion at all: it's like jumping into an argument about sex before marriage or ethical non-monogamy if you believe all sex is immoral—you cannot make a constructive contribution. (You're also calling pretty much everyone else in the thread unethical and dishonorable.)

I think I put enough caveats in my “take” to cover yours.

We have different opinions, and that’s fine by me. But I’m not talking about “professionally published authors;” I covered that too. I’m talking about writers on Lit.

 

[Throwaway256897q6r72]

A thought occurred to me this morning. I'm not a professional writer. I know the basics of English grammar, but do get confused on when to put a comma. Like that last sentence. Did it need a comma before but? Should I have used a semi-colon between sentence and did? Those are the type of things I use an online editor for. And it catches the occasional miss spelled word.

How is this any different if I had wrote my story up in a program such as Microsoft Word and used the spell checker on there?

 

[nice90sguy]

The differnce is that the spell-checker corrects (conforms to the generally-accepted) your spelling, and a grammar checker "corrects" your grammar -- conforms it to a generally-accepted standard of grammar.

An AI assistant goes even further, conforming your writing to more general readability rules (which it does NOT learn explicitly).

Spell-checkers are considered OK (I doubt if a story would be rejected purely becuase it didn't find any spelling errors). But, with with grammar, the rules are much more flexible. The reason you "get confused on when to put a comma", is that the rules aren't hard and fast -- except in legal. professional or scientific writing, for instance. In fiction, it's precisiely because the rules are flexible that writing which appears too conformant to an AI assistant's "take" on correct grammar will get flagged.

 

[Throwaway256897q6r72]

Thank you, that makes perfect sense. Now, just to wait to hear from Laurel if my stories will be republished or not.

 

[Voboy]

A thought occurred to me this morning. I'm not a professional writer. I know the basics of English grammar, but do get confused on when to put a comma. Like that last sentence. Did it need a comma before but? Should I have used a semi-colon between sentence and did? Those are the type of things I use an online editor for. And it catches the occasional miss spelled word.

How is this any different if I had wrote my story up in a program such as Microsoft Word and used the spell checker on there?

The best thing about English is that its rules are flexible, making it easy to convey shades of meaning. I can think of situations where you wouldn’t include a comma before “but” and it would connote a specific mood. Or? Include it. Either is okay. Same with your other example.

You don’t really need much editing! If it looks good to you and conveys the meaning you seek, others will like it too.

 

[MillieDynamite]

I'm so pleased someone thinks my writing is better than, "Okay!"

Thanks, hon. I appreciate that. Your stuff's great.

 

[NoTalentHack]

Welp, got the story in. You know what ended up being as big a help as anything? Having Word read the doc to me. Let me trim some stuff and/or rework things that worked better in my head than on the page, plus catch a couple of dropped or extraneous articles.

 

[AwkwardlySet]

I'm so pleased someone thinks my writing is better than, "Okay!"

Bah, leave it to NTH to miss an easy zinger. "Err... yeah, your writing is nice too!"

 

[Bramblethorn]

Actually, the logical inference from an excerpt being diagnosed differently by different detectors should be that no one might be safe in assuming they won't be tagged. However, adopting the right stance never hurts.

And "no one might be safe" implies "you might not be safe", hence the latter is also a logical inference. It's possible for several things to be true simultaneously!