Why is everything labeled as AI now????? I can't even post a story anymore.

J'ai dû utiliser Google Traduction pour le lire, et à ce stade, on voyait clairement les choix de mots, la structure des phrases et le ton d'un juriste. Je ne prétends pas que vous ayez fait de même pour votre texte original, je n'en suis pas capable, mais ce forum n'est pas le lieu pour porter ce genre de jugement.

J'ai bien aimé ce que j'ai lu. Je suis désolé que votre texte ait été refusé (en supposant qu'il s'agisse bien du texte que vous avez soumis).
JJe confirme que c'est une partie du texte présenté.. merci au moins pour votre appréciation et Google traduction fonctionne pas si mal si le texte initial est correcte... alors comment trouver de l”IA?
 
Anyone with even the most passing familiarity with Australian swans would know that any finite number of examples of a thing that has a trait is not in fact meaningful evidence that ALL examples of that kind of thing have that trait.

No number of white swans demonstrated is dispositive evidence that black swans do not exist. Similarly, no number of white shoes is dispositive evidence that all crows are black.

Universal claims have a VERY high philosophical bar to jump before they can even be seriously considered.
 
You said all AI detetors use AI. Do you stand by that assertion?
He doesn't have to stand by it. I can disprove it right here and now:

Python:
def is_text_ai_generated(text, threshold=0.3):
    word_count = len(text.strip().split())
    mdash_count = len([_ for c in text if c == '—'])
    mdashes_per_word_pct = mdash_count * 100 / word_count
    return mdashes_per_word_pct >= threshold

There you go, an AI detector, using the oft-cited heuristic about em-dashes.

Is it a good detector? Hell no. But it exists, and it sure as hell does not use AI itself.
 
Vu la mise en forme, je dirais IA. Au début, j'ai cru que votre message était la version française du Lorem Ipsum.

Je déteste les tirets, mais ne me citez pas là-dessus.
He bien vous avez tort, ma seule intelligence utilisé est humaine.. je ne crois pas être un robot😡
 
JJe confirme que c'est une partie du texte présenté.. merci au moins pour votre appréciation et Google traduction fonctionne pas si mal si le texte initial est correcte... alors comment trouver de l”IA?
J'aimerais vraiment, vraiment pouvoir faire plus. Désolé. Ce que je peux suggérer au-delà de la resoumission avec une note n'est pas utile en langues autres que l'anglais, et je ne peux pas expliquer le mécanisme sans expliquer comment cela fonctionne. Le détecteur d'IA de Lit's ne fonctionne que dans le vide, comme une boîte noire qui n'est pas comprise. Si c'était le cas, ce serait contournable et le site serait inondé de boue au point qu'aucun de nous ne pourrait attirer l'attention sur notre travail.
 
Anyone with even the most passing familiarity with Australian swans would know that any finite number of examples of a thing that has a trait is not in fact meaningful evidence that ALL examples of that kind of thing have that trait.

No number of white swans demonstrated is dispositive evidence that black swans do not exist. Similarly, no number of white shoes is dispositive evidence that all crows are black.

Universal claims have a VERY high philosophical bar to jump before they can even be seriously considered.
I *think* you're getting at false positives, and you're right. It's entirely possible. That has always been the most unfortunate aspect of Lit's implementation of an AI detector. It's very aggressive, and only knows how to spot what it knows how to spot. If you tell it no orange, it'll stop oranges and foxes and some very ripe wheels of Wisconsin Cheddar despite the fact that these are very different things. You can fine tune the shade, within the context of this metaphor, but it has limitations.

(or I completely misunderstood you and you're talking about Metabob's broad claims of AI to catch AI)
 
(or I completely misunderstood you and you're talking about Metabob's broad claims of AI to catch AI)
I was in fact talking about Metabob's claims. He showed some AI-using AI-detectors and claimed that meant that all AI-detectors used AI. And that's just absolutely not how logic works.
 
The case for the prosecution rests.

That’s what you said and its demonstrably false. Admit that some don’t (and in particular that we have no fucking clue about what Literotica does) and we’re cool.

Keep arguing without addressing what you actually claimed and I see no point in this convo.
Can't spend any more time on this, sorry. I finally submitted a chapter I've been finalizing for weeks in between all this noise plus dishes, laundry, and bathroom cleanup, then dishes to prepare for a dinner party I'm hosting, then guests arrive with more wine. Tomorrow, jamming at the bar (with beer), followed by foobah and more beer. Maybe Monday.

Stepping way from the computer now; it's being co-opted into use for the higher purpose of being a jukebox.
 
Last edited:
Can't spend any more time on this, sorry.
That’s secret code for, “You might have had a point, Emily. But I’m not going to admit it.”

Enjoy your dinner party 😊. I’ll accept your tacit concession. Moving on…
 
That’s secret code for, “You might have had a point, Emily. But I’m not going to admit it.”

Enjoy your dinner party 😊. I’ll accept your tacit concession. Moving on…
Oy, more nonsense. My chowder is the main dish. First task: chopping taters.
 
Oy, another "because I said so" argument. Provide examples, please.

Here are some from my side. They all use AI and they all say as much:

https://gptzero.me/
https://www.grammarly.com/ai-detector
https://ai-detector.net/
https://originality.ai/

Here, these and others are compared:
https://nypost.com/shopping/best-ai-detectors/

I have tried several. I thought that Grammarly's sucked.
I mean, you're "not-wrong" in that, yes, those are all AI AI detectors.

There are detectors which are not themselves AI, though. There is such a thing. Some of them work through "fingerprinting." There's a subject for you to google. I don't know whether that's how Lit is doing it, but fingerprinting is designed to make (at least some) AI output detectable with non-AI testing.
 
I mean, you're "not-wrong" in that, yes, those are all AI AI detectors.

There are detectors which are not themselves AI, though. There is such a thing. Some of them work through "fingerprinting." There's a subject for you to google. I don't know whether that's how Lit is doing it, but fingerprinting is designed to make (at least some) AI output detectable with non-AI testing.
Dinner party over. My chowder and other things went over very well, among distinguished company.

I will look up this "fingerprinting" technique you mention. I'm pretty ... discerning as a software developer, so I might have ... insights. But honestly, this is the first potentially verifiable clue that anyone else on this thread has offered other than 'because I said so.' I appreciate that. It is probably not negligible, but after a bunch of good wine with a bunch of people who liked my chowder (maybe someone might know which kind if they read my latest story, which is pending, though I doubt that anyone here will do so), I am not reliable as a source for additional text tonight.

Still trying to process that 30+-year-old cab I brought out at the end of dinner. I have a tiny bit left in my glass, next to my laptop after disconnecting it as my dinner-party jukebox.

I might have the luxury of more time to respond in more detail on Monday. Tomorrow: afternoon jamming and beer followed by foobah with more beer. I will try to get more writing done in the morning.

Best wishes to all.
 
Last edited:
Time enough to reply, I see 🤣
I misspoke. Chopping taters was not my first task after stepping away from my laptop for dinner prep. The first thing was getting my mixing board set up with the right cables between my laptop and receiver. It took a little while before I got it right; I have MANY audio cables that can be set up in multiple configs. There was also the bacon thing and the chopping onions and the whole "I don't want my fingers that are covered in wet onion juice going near my laptop" thing. Only then did I chop the taters and dice the garlic and the parsley. I put one thing after another on the heat. One pot, because of course. No leftovers. Yum.
 
He doesn't have to stand by it. I can disprove it right here and now:

Python:
def is_text_ai_generated(text, threshold=0.3):
    word_count = len(text.strip().split())
    mdash_count = len([_ for c in text if c == '—'])
    mdashes_per_word_pct = mdash_count * 100 / word_count
    return mdashes_per_word_pct >= threshold

There you go, an AI detector, using the oft-cited heuristic about em-dashes.

Is it a good detector? Hell no. But it exists, and it sure as hell does not use AI itself.
Irrelevant. Lit doesn't use this and never will, because your code will never function to their needs, as you freely admit. Anyone, even a competent coder as you seem to be, can produce shit code. Just as an AI can. Congrats to you for doing so as if it matters.

I use em-dashes, fwiw. I have cited specific examples of how I have used them here, and that I thought their prevalence was irrelevant to whether an AI generated their use, or didn't. My publisher doesn't like them but lets me use them, because em-dashes are not wrong. I don't think their usage indicates anything other than preference, certainly not whether they were generated by AI. Laurel doesn't seem to like them either but has tolerated mine, at least until now. Maybe that'll change after all the noise here. Who knows.
 
Last edited:
Time enough to reply, I see 🤣
OK, since this seems to be what you want and you have repeatedly invited it even though I recognize that responding to this sort of noise is a waste of my time, let’s do this.

First, the subject of this thread: Why is everything labeled as AI now????? I can't even post a story anymore. Sounds like a cry for help to me. How 'bout you?

1. So I helped, describing how an AI can be used to rewrite a human-generated story and providing a concrete example comparing the original text and the AI-modified rewrite so that examining the difference between the original and the rewrite would be helpful, because helping is a thing that I do:
https://forum.literotica.com/thread...n-post-a-story-anymore.1648707/post-102268682 .

2. As someone who has built a clean-room working AI from scratch (can anyone else who frequents this forum say the same?), I provided details on how they work, and why what happens inside an AI is very difficult to map to human decision-making process. Not complete details, of course, but the bones of an outline understandable by laypeople (like you), plus a link that could help people learn more, because helping is a thing that I do:
https://forum.literotica.com/thread...-story-anymore.1648707/page-11#post-102274830

3. I asserted that this site’s AI detector is itself an AI, because that’s how most AI detectors work. This might have been incorrect, but no evidence to the contrary has yet been offered beyond 'because I said so':
https://forum.literotica.com/thread...n-post-a-story-anymore.1648707/post-102275068

4. I presented a publicly accessible website (ChatGPT, which is an AI) showing how one could A) submit human-written text and see how this AI detector evaluates it, B) rewrite that original text using AI, and C) submit an AI-rewritten version and see how this AI detector evaluates it. A method that can be used to produce replicable results. Y’know, like science:
https://forum.literotica.com/thread...-story-anymore.1648707/page-12#post-102275280

5. I agreed with @AwkwardlySet that Lit’s approval mechanism might include a whitelist, something that has been discussed on this forum many times before and backed up by my own (anecdotal) experience here:
https://forum.literotica.com/thread...-story-anymore.1648707/page-12#post-102275395

6. I have challenged multiple posters to back their assertions with verifiable examples. Exactly none have done so, including you:
https://forum.literotica.com/thread...n-post-a-story-anymore.1648707/post-102275485

7. I provided multiple examples of AI detectors that are easily accessible on the web, and asserted that they all use AI and state as much, because helping is a thing that I do:
https://forum.literotica.com/thread...n-post-a-story-anymore.1648707/post-102275933

8. I summarized what I had previously contributed on the subject here, because helping is a thing that I do:
https://forum.literotica.com/thread...n-post-a-story-anymore.1648707/post-102276222

9. Stepped away from the discussion because I had a dinner party to prep for and needed to convert my laptop from a board contributor to a jukebox:
https://forum.literotica.com/thread...-story-anymore.1648707/page-14#post-102276540

10. After configuring the music setup, I shot down someone else’s irrelevant example of admittedly useless code that can’t be used by Lit:
https://forum.literotica.com/thread...n-post-a-story-anymore.1648707/post-102277940

By contrast, you have:
A. Asserted or that I am wrong (or ad-hominem implied that I am clueless, including with a meme) while providing no evidence for that assertion:
a. https://forum.literotica.com/thread...n-post-a-story-anymore.1648707/post-102275748
b. https://forum.literotica.com/thread...-story-anymore.1648707/page-13#post-102275902
c. https://forum.literotica.com/thread...-story-anymore.1648707/page-13#post-102276023
d. https://forum.literotica.com/thread...-story-anymore.1648707/page-13#post-102276030
e. https://forum.literotica.com/thread...-story-anymore.1648707/page-13#post-102276292
f. https://forum.literotica.com/thread...-story-anymore.1648707/page-13#post-102276300

B. Claimed 'victory' by my inability to respond (due to needing to prepare for and host said dinner party) for several hours. There’s probably a technical term for this sort of intentional rhetorical misdirection, but I can’t find my Rhetoric book ( https://www.goodreads.com/book/show/28260478-trivium ) just now. Perhaps that’ll be an additional reason for you to claim 'victory' again:
g. “The case for the prosecution rests”: https://forum.literotica.com/thread...n-post-a-story-anymore.1648707/post-102276364
h. “I’ll accept your tacit concession”: https://forum.literotica.com/thread...n-post-a-story-anymore.1648707/post-102276562
i. “Time enough to reply, I see”: https://forum.literotica.com/thread...n-post-a-story-anymore.1648707/post-102276703

The best that could be said about your responses to me on this thread is that they are unsupported nitpicking. One could go down the list of descriptors from there.

I was warned by multiple people more than two years ago that engaging with you would be a mistake. That was the most stressful time of my life so I made the strategic decision (recommended in those private discussions) to put you on my ignore list because it would be pointless and aggravating to engage further. That stressful time in my life is now past, but after a year or so I deleted that disengagement, mostly because I don’t like ignoring people. However, it seems clear that going forward, ignoring you again needs to be the strategy. I’m curious to see if you have more to say before I engage that ignore. I’m guessing that there will be a tale told ..., full of sound and fury, signifying nothing. I challenge you to do better.

It is not worth my time or life investment to continue engaging with someone who has nothing to contribute but attention-seeking and insults. Good day and good riddance.

Addendum: This post might be deleted and/or result in getting me banned but I care little about either of those potentialities, though I would regret leaving Literotica and its writer community that has been mostly supportive to my writing career. With a few notable exceptions.

There's three hours of my life gone forever. Congrats.
 
Last edited:
Your words:
"Keep arguing without addressing what you actually claimed and I see no point in this convo."
https://forum.literotica.com/thread...n-post-a-story-anymore.1648707/post-102276364

"That’s secret code for, “You might have had a point, .... But I’m not going to admit it.”
... I’ll accept your tacit concession. Moving on…"
https://forum.literotica.com/thread...n-post-a-story-anymore.1648707/post-102276562

"Time enough to reply, I see 🤣"
https://forum.literotica.com/thread...n-post-a-story-anymore.1648707/post-102276703

I challenged you to do better. Can you?
 
Your words:
"Keep arguing without addressing what you actually claimed and I see no point in this convo."
https://forum.literotica.com/thread...n-post-a-story-anymore.1648707/post-102276364

"That’s secret code for, “You might have had a point, .... But I’m not going to admit it.”
... I’ll accept your tacit concession. Moving on…"
https://forum.literotica.com/thread...n-post-a-story-anymore.1648707/post-102276562

"Time enough to reply, I see 🤣"
https://forum.literotica.com/thread...n-post-a-story-anymore.1648707/post-102276703

I challenged you to do better. Can you?
I’m deeply impressed that I got so much under your skin by simply pointing out the meaning of the word, ‘all.’

I hope you enjoyed writing that, as I have no intention of reading a word. But well played, please feel free to count this as a victory. Maybe someone else will care enough to read it and tell you what a great job you did.

What is life without hope, right?
 
I’m deeply impressed that I got so much under your skin by simply pointing out the meaning of the word, ‘all.’

I hope you enjoyed writing that, as I have no intention of reading a word. But well played, please feel free to count this as a victory. Maybe someone else will care enough to read it and tell you what a great job you did.

What is life without hope, right?
Life is life. No projecting on others necessary.

It isn't that you got under my skin, it is that you challenged my writing and my expertise when you clearly knew nothing about what you were writing. Now you're getting called on it, and ... again, your words:

"Keep arguing without addressing what you actually claimed and I see no point in this convo."
https://forum.literotica.com/thread...n-post-a-story-anymore.1648707/post-102276364

In turn, I challenged you to do better, and you are admitting that you won't, which means you can't. I welcome you to actually try rather than just continue to snipe. I don't care about "victory" -- you claiming my "tacit concession" without justification is your thing, not mine. I'm happy enough to debunk your specious, unsubstantiated claims, give you the opportunity to respond, and move on if you can't. Or won't.

Do better.
 
Last edited:
I mean, you're "not-wrong" in that, yes, those are all AI AI detectors.

There are detectors which are not themselves AI, though. There is such a thing. Some of them work through "fingerprinting." There's a subject for you to google. I don't know whether that's how Lit is doing it, but fingerprinting is designed to make (at least some) AI output detectable with non-AI testing.
I looked this up. It echoes a conversation I have with my other tech friends on a fairly regular basis. It seems that "fingerprinting" relies on voluntary cooperation by outfits that can do generative AI to include indicators like checksums that are not immediately obvious.

That's a pretty good way to weed out legit outfits who have AIs that generate text, and Lit would be foolish not to take advantage, but for people like me who have developed their own clean-room AI, or for other domestic or foreign outfits with more capability than my framework ever had that do not wish to voluntarily conform to including these "fingerprints", it won't work as well.
 
Back
Top