First AI Rejection Today - Machine Translation

I understand there probably aren't too many users publishing translations of their stories, which might explain why it hasn't been dealt with before. It really does need to be addressed, because the notion that machine translations of original human-written stories should count as "AI-generated" is absurd. It would be like saying that a human translator of a story should be credited as the "original" author of the story instead of for translating someone else's work, entitling them to the copyright and the royalties.

Translators do get copyright in their translations. If I were to publish a copy of Seamus Heaney's or Maria Dahvana Headley's new translations of "Beowulf" without authorisation, I could expect to hear from their lawyers.

Where the original is still under copyright, a translation is considered a derivative work, which essentially means that the original author and the translator both have rights to it and any publication/copying needs to be authorised by both of them.

Very often the translator is working in a "work for hire" arrangement, in which their contract will state that they sign over their rights to the translation in exchange for their payment. But unless that happens, the original author cannot legally publish that particular translation.

Given that the translator is effectively considered a second author for legal purposes, it makes sense that a site that forbids AI-written stories would also forbid AI-translated stories.

https://blogs.loc.gov/copyright/2022/10/copyright-in-translation-gregory-rabassa/
 
Already done, fingers crossed.

I grant, knowing a policy or clarity around this “special case” would be worthwhile, but I kind of feel like the outcome of that should have been obtained first, before all the outrage.
 
I agree with you that the quality of the translation often leaves something to be desired, but the issue isn't the quality of the translation; the issue is the story being rejected on the grounds that it's "AI generated". I've published twenty other machine translations here on Literotica, and that didn't stop Laurel and whatever filtering software she uses from approving each of them. Either, this is the first time the software has noticed (which it shouldn't, because the original English version was written by me unaided), or this is the false positive. Either way, there's nothing in Literotica's AI policy about machine translations, which rely on AI but don't generate any original content.
I stand corrected and would offer this suggestion.

I've had a couple stories rejected for political content, though neither story advocated for any ideology or political position. I've since gotten a couple similar stories published with no problems by adding a "note to admin" explaining that while politics are a background for the story, they are only that. Try adding a "note to admin" telling Laurel that the story is a machine translation of a story you authored and include a link to that story.
 
Transformer architectures, created to translate from one language to another, are the basis of LLM's. In most Copyright jurisdictions the user of the program is the copyright holder; in the USA the Copyright Office is 'confused'. There, the program cannot hold the copyright, nor can the user, because the user does not pass the creative act test.

In translating from one language to another, the programs have the same problems as translating from English to English, which is what they're most commonly used for. Check out some of the snippet threads and you'll see how one concatenation of words can be misinterpreted by both native English speakers and LLMs, if shorn of sufficient context.

This may go some way to explaining the uncertainty and inconsistency which a site like Lit, which is a non-AI site, shares with the rest of the world.
 
Thanks for taking the time to give your thoughts on this.
I suspect that Lit has always known you were submitting machine-generated translations, and that this rejection of one instance is a slap on the wrist to get your attention. I see your catalog has some stories with French AND German translated repostings. I don’t remember what the context was, but I do remember Laurel telling me once that the site doesn't want to host redundant content.** Machine generated translations are toe-ing that line.
So Lit has always had a problem with my machine-generated translations but only now decided to make it known to me? That's practically the definition of arbitrary enforcement, especially since the AI policy has nothing to say about translated works either way. As for redundant content, judging by who has commented on my English, French, and German stories, I don't see any overlap between the different readerships, so I don't think fears of redundant content apply here. At least, they shouldn't.
I can also tell you that what you're doing violates the TOS for the Kindle store. Machine translations of content will get your account sent to the shadow realm. I appreciate that Lit is a different site with its own rules, but I think it's worth noting when a significant player in the field takes a stance like that (and for context that rule has been in place since at least 2020) At any given moment, Lit can simply decide that "If a rule is good enough for Amazon, it's good enough for us."
I can tell you for a fact that's NOT against Kindle's TOS. All you have to do is declare honestly on the Kindle eBook Content page whether AI was used to generate the cover, the content, or the translation, and specify how much manual editing you performed on the AI-generated content. Amazon knows I've submitted machine translations of my English works for sale because I've selected the option declaring it upfront.
EDIT:** In other words, if the content of your German language submission is what a user would get if they simply ran your existing story through Google Translate, then the site would prefer readers do that themselves and only host your original work in the language you originally wrote it.
That would seem to defeat the point of having different language versions of the site. Why bother allowing non-English speakers to browse the site in their own language when we can force them to use Google Translate or some other program to search for an English story they might like and then read it?
EDIT2: I did not make my second point clearly. Perhaps this is not a case of "oh here's this blameless thing why do they suddenly have a problem with it?", but rather "This was always a problem, but they let it go since you are a consistent (authentic) author. Now it seems like it's becoming a pattern and they're not okay with that."
Clarity, or lack thereof, is still my main problem with this. If Lit is fine with machine translations by an author within certain limits, they should say so. If they're not fine with machine translations creating "redundant" content, they should also say so.
 
I wrote up a long response, but ended up deleting it because it doesn't seem like you want direction. It seems like you want sympathy because you think your version of breaking the rules is "different."

Your non-English content definitively violates the guidelines, but the guidelines also state that Lit's policy on AI is evolving. Good luck getting on the right side of that.
 
Words like "genie" and "bottle" come to mind.

Welcome to a world where nothing is believable, a world where the victory of gushing nerddom is complete.
 
I wrote up a long response, but ended up deleting it because it doesn't seem like you want direction. It seems like you want sympathy because you think your version of breaking the rules is "different."

Your non-English content definitively violates the guidelines, but the guidelines also state that Lit's policy on AI is evolving. Good luck getting on the right side of that.
With all due respect, twenty approved submissions followed by a rejection IS different. Suddenly, the filtering software has noticed that the translation was produced by a computer program. That's a successful detection rate of 4.67%, which makes a joke out of any policy on the topic, however well intentioned.
 
From: https://www.literotica.com/faq/publishing/publishing-ai

One area of particular concern is software or apps that “rewrite” your paragraphs or stories for you. The text that results from rewriting features is definitely AI generated. These type of apps are replacing original human written text with generic AI generated text, and should be avoided when submitting work to Literotica. Readers prefer to experience your worlds and your fantasies in your own unique voice, rather than having them smoothed into a generic artificial voice available to everyone else using the same software.

(Bolded section by me, for emphasis) This is what you did. It violates the guidelines. That you got away with it is not the win you think it is.
 
Last edited:
From: https://www.literotica.com/faq/publishing/publishing-ai

(Bolded section by me, for emphasis) This is what you did. It violates the guidelines. That you got away with it is not the win you think it is.
This refers to writing aid programs that alter the tone and phrasing of text in the same language and has nothing to do with faithfully translating a text from one language to another. If the translation program "rewrites" the text in this way, then it's not an accurate rendering of the original and is not a translation. They are two entirely different types of programs, so your quote is irrelevant, and your claim that I'm trying to "get away with" violating the guidelines is junk.

If Lit wants to include machine translations under its AI policy guidelines, that's fine, it should do so to settle the matter once and for all, and I'll abide by that; but don't gaslight me by claiming the guidelines say something they clearly do not say.
 
This refers to writing aid programs that alter the tone and phrasing of text in the same language and has nothing to do with faithfully translating a text from one language to another.

But "faithfully translating" between languages isn't really a thing, because languages don't map to one another one-to-one.

Even for two languages as closely related as English and German, a single sentence like "On the other side of the wall, you can see a star" has at least forty possible translations which differ significantly in meaning. To translate that properly, one needs to understand the meaning of the sentence, which requires understanding how it fits into the surrounding text.

On the other hand, if I have a character talking about how the Swedish Navy have started putting bar codes on their ships so they can scan the navy in, there is no satisfactory way to translate that into German (or most other languages). You can keep the literal meaning or you can have the character making a terrible pun, but you can't do both. Probably the best "translation" here would be to ditch literal translation entirely and replace it with some equally bad pun in German.

A machine translator which can do that kind of thing...is hard to find, but if it exists, then it's making the same kind of judgements that are involved in rewriting an English sentence, which probably requires it to be trained on translations performed by humans who were making such judgements, which raises the same kind of ownership issues.
 
Back
Top