Update content warnings to include AI usage

[Theresaisnotmenhera]

As per Scribble Hub's content guidelines, stories that consist primarily of AI-generated content are prohibited, although the use of AI is not strictly banned. The pertinent rule states:

AI Generated Stories - Stories created mostly by AI will be rejected. You can use AI to help create your stories but most of the story should still be written by you.

However, the language used in this guideline is vague in defining what constitutes "mostly." This has led to the acceptance of several stories with a significant amount of paraphrased content or portion generated by AI, carrying clear voice and tones of AI rather than of a human writer. In light of the inability to filter these stories, it is suggested that a model similar to that of Royal Road be adopted. Royal Road currently employs two tags:

AI-Assisted: The author has used an AI tool for editing or proofreading. The story thus reflects the author’s creativity and structure, but it may use the AI’s voice and tone. There may be some negligible amount of snippets generated by AI.


AI-Generated: The story was generated using an AI tool; the author prompted and directed the process, and edited the result.

Regardless of whether the two are to be merged into one, a more concise language and updated content warnings would provide readers with necessary information and a greater ability to filter out unwanted material, including establishing a clearer policy for the site.

Note: Content warnings should not be needed if AI is a glorified thesaurus for you, for example.

 

[tiaf]

The Buggywhip manufacturers and slave plantation masters salute you. Stay the course. AI is as evil as electricity, hygiene, and the automobile combined!

AI-written content is crap. It always has been crap and it always will be crap. Guess what? 95% of DeviantArt 'art' was crap long before AI art was a thing, as was 95% of all written stories. If anything, AI has RAISED the quality of these uncreative brutes.

We don't need tags, what we need is for people who stumble across bad content to mark it as bad content. Why? no one is stupid enough to label their content as garbage, even if it is. 'recognizing' AI styles is one thing, but I have also read handwritten shlock that is every bit as poorly styled as AI junk.

And honestly, I love the way that AI artwork is clearing out all the complete nimrods from visual art with no talent that think they deserve to get paid like the next Boris Vallejo.
If your 'art' is poor enough that it can be replaced by AI, it deserves to be.

I still prefer human crap to AI crap. There is the difference between what kind of crap humans can create and sometimes it just gets worse than I could imagine, but then, once in a red moon, there is one unpolished crap that can turn into a shiny gem.

AI is like my youtube recs, always the same repetition.

 

[John_Owl]

The Buggywhip manufacturers and slave plantation masters salute you. Stay the course. AI is as evil as electricity, hygiene, and the automobile combined!

AI-written content is crap. It always has been crap and it always will be crap. Guess what? 95% of DeviantArt 'art' was crap long before AI art was a thing, as was 95% of all written stories. If anything, AI has RAISED the quality of these uncreative brutes.

I do agree. And this from someone that uses AI as a sounding board for my stories (they're written by me, but I use it to search for options I may not see myself), I remember pre-AI, when I was reading a fanfic about RWBY that felt like it was written by a preteen neckbeard with one hand in his pants (broken syntax, grammar that read like someone used google translate between five different languages then copy-pasted, and trying to make it 'sexy' while still keeping it acceptable). THAT is the measure against which all bad writing is measured for me.

I've never read an AI story that bad. they're still crap, yes. But not NEARLY that bad. stories like that, written by humans, are worse than what AI will make. But at the same time, no AI will be able to write shakespeare or tolkien. ALL AI falls somewhere in the middle, while Humans run the full gamut, though most of us fall more towards the measure of 'bad' honestly.

All this to say this: AI, while still a mud pit, has raised the quality of the mud from pig-slop.

 

[Senruika]

And how do you guys would know wether a story is AI generated or not?

 

[Dec]

And how do you guys would know wether a story is AI generated or not?

It's super easy to spot one, actually.
Everything is flat, no emotions, overly fancy wording that has nothing to do with the situation at hand, broken logic, and so on.
You instantly know that it was made by AI when you read that text.

 

[beast_regards]

How do you define the AI?

The term has been used, and overused, to the extent it lost its original meaning, to the point you could replace it with "Banana Inserter" and no meaning would be lost. Everything has "AI" features, even my coffeemaker, not to mention every piece of software marketed in the last few years, regardless if it is true or not. The operation system is supposedly "AI", as is the Google browser spellchecker, Google Docs, the Pro-Writing-Aid, the Grammarly, the Thesaurus, my coffeemaker, everything. It, of course, doesn't stop anyone from accusing each other of "using the AI" because, well, it has no meaning and you don't even need a proof, so why not?

What it would the tag even do?

Would it hide the story or merely hang there, warning the readers?

Is it a tag that the users would enable manually, or would it be automatically assigned by some script running the text through the "detector" software?

If it is manual, why would anyone, in the right might, enable the "bully enabler, down vote me here", especially if no one could agree on what it means?

And if it done through automatic process, how do you handle the false positives from the software? Run any older novel, traditionally published, older than 5 years, and you would still get some of them flagged as AI as the software couldn't tell the difference.

How you could even prove you aren't using AI? Recording? A witness?

 

[StoneInky]

As per Scribble Hub's content guidelines, stories that consist primarily of AI-generated content are prohibited, although the use of AI is not strictly banned. The pertinent rule states:



However, the language used in this guideline is vague in defining what constitutes "mostly." This has led to the acceptance of several stories with a significant amount of paraphrased content or portion generated by AI, carrying clear voice and tones of AI rather than of a human writer. In light of the inability to filter these stories, it is suggested that a model similar to that of Royal Road be adopted. Royal Road currently employs two tags:



Regardless of whether the two are to be merged into one, a more concise language and updated content warnings would provide readers with necessary information and a greater ability to filter out unwanted material, including establishing a clearer policy for the site.

Note: Content warnings should not be needed if AI is a glorified thesaurus for you, for example.

I also agree that we need this. Another for.

Plus, I hope rules will be implemented against AI assisted novels asking for reviews in the Feedback forum.

How do you define the AI?

The term has been used, and overused, to the extent it lost its original meaning, to the point you could replace it with "Banana Inserter" and no meaning would be lost. Everything has "AI" features, even my coffeemaker, not to mention every piece of software marketed in the last few years, regardless if it is true or not. The operation system is supposedly "AI", as is the Google browser spellchecker, Google Docs, the Pro-Writing-Aid, the Grammarly, the Thesaurus, my coffeemaker, everything. It, of course, doesn't stop anyone from accusing each other of "using the AI" because, well, it has no meaning and you don't even need a proof, so why not?

What it would the tag even do?

Would it hide the story or merely hang there, warning the readers?

Is it a tag that the users would enable manually, or would it be automatically assigned by some script running the text through the "detector" software?

If it is manual, why would anyone, in the right might, enable the "bully enabler, down vote me here", especially if no one could agree on what it means?

And if it done through automatic process, how do you handle the false positives from the software? Run any older novel, traditionally published, older than 5 years, and you would still get some of them flagged as AI as the software couldn't tell the difference.

How you could even prove you aren't using AI? Recording? A witness?

Bringing older unedited versions of your writing? Google doc saves every edit and when it was made. Even Scribblehub saves a list of how many edits were made. A writer who polishes their work to the point it gets mistaken for AI...would be a very strict editor. They would have tons of edits. Screenshots would suffice as proof.

And even if it's not enforced strictly, I hope people still have the option to mark their novels as AI edited. It's not a crime, it's just something people might want to know. And new writers might forget or not realize.

 

[beast_regards]

Bringing older unedited versions of your writing? Google doc saves every edit and when it was made. Even Scribblehub saves a list of how many edits were made. A writer who polishes their work to the point it gets mistaken for AI...would be a very strict editor. They would have tons of edits. Screenshots would suffice as proof.

And even if it's not enforced strictly, I hope people still have the option to mark their novels as AI edited. It's not a crime, it's just something people might want to know. And new writers might forget or not realize.

The Google Docs saves suggestions...

If I brought up the screenshot of suggestions, it would display not only my name but also the name of the person who volunteered to proof-read my story. There were quite a few, in fact, three people so far helped, each logging with their google account, two of which use what I assume are their real names. Perhaps they use made up names in Google account, that's a possibility too, but even then, attempting to doxx them doesn't feel like gratitude for helping me.

And all of that to prove the "AI" (which isn't even the real AI) wasn't involved in making the story that is, despite all its flaws, a free to read?

An argument could be me against selling the randomly generated text, as omitting it was a generated by the "AI" would be misleading the customer about the product composition, and there are laws against that, but free content?

How does inciting the doxxing witch hunt help anyone?

Who is even the arbiter, the judge, of this "AI proof" scheme?

 

[StoneInky]

The Google Docs saves suggestions...

If I brought up the screenshot of suggestions, it would display not only my name but also the name of the person who volunteered to proof-read my story. There were quite a few, in fact, three people so far helped, each logging with their google account, two of which use what I assume are their real names. Perhaps they use made up names in Google account, that's a possibility too, but even then, attempting to doxx them doesn't feel like gratitude for helping me.

And all of that to prove the "AI" (which isn't even the real AI) wasn't involved in making the story that is, despite all its flaws, a free to read?

An argument could be me against selling the randomly generated text, as omitting it was a generated by the "AI" would be misleading the customer about the product composition, and there are laws against that, but free content?

How does inciting the doxxing witch hunt help anyone?

Who is even the arbiter, the judge, of this "AI proof" scheme?

I meant the people who run the site? You wouldn't doxx it publicly, of course. But if your story gets flagged for not having the AI-edited tag, except you didn't use AI, and you need proof, this can be said proof you give to the moderators. Access to the doc and history, or screenshots of said history.

Even if it's free to read, the site is also freely hosting your novel. If you don't want to post cuz you don't want to mark AI edited stories, just don't post. Simple.

And as I said, even if the tagging/marking rule is not strictly enforced, I still hope people get the OPTION to mark their stories with the mark or tag. Right now, newbie author may not even know that's a thing. That's all.

 

[CountVampyre]

AI is not the problem, but the user. Just like with a knife, you can use it for good, like cooking food for your friends, or injure them. It's the same with AI. You can use it to increase your quality through spell-checking and similar tools. Then again, there is the dark path. AI brings opportunities for quick cash. With a little bit of tropes here, popular genres there, you can create something quite simple. Be it on RR, SH, or WN, those stories are popular. Of course, 'authors' of these need to invest time in their works too, probably market them and so on, but that wasn't what I wanted to point out either. I just think we should not judge AI as a whole. As for a warning? I think it's fair that one needs to issue them as soon as a clear undertone of AI can be seen.

 

[beast_regards]

I meant the people who run the site? You wouldn't doxx it publicly, of course. But if your story gets flagged for not having the AI-edited tag, except you didn't use AI, and you need proof, this can be said proof you give to the moderators. Access to the doc and history, or screenshots of said history.

Even if it's free to read, the site is also freely hosting your novel. If you don't want to post cuz you don't want to mark AI edited stories, just don't post. Simple.

And as I said, even if the tagging/marking rule is not strictly enforced, I still hope people get the OPTION to mark their stories with the mark or tag. Right now, newbie author may not even know that's a thing. That's all.

Flagged by who?

By the random user?

Wouldn't it create a perfect opportunity for sabotage someone's else novel?

Flag it, and force the author to either doxx himself, or delete their stories, or put their story in some form of limbo until the overburdened moderation replies while the rumor mill does its work?

Does the AI generated art I post be used as the argument against me?

See, there was the case among the artists (i.e. people who draw pictures) where they falsely accuse other using AI as the way to tarnish someone's reputation. They are much more sensitive than this site is, and we I doubt we need to bring their witch hunts here.

Just go to the Royal Road and look what toxic hellhole in turned into when one allows writers to sabotage their competition for sake of profit.

 

[StoneInky]

Flagged by who?

By the random user?

Wouldn't it create a perfect opportunity for sabotage someone's else novel?

Flag it, and force the author to either doxx himself, or delete their stories, or put their story in some form of limbo until the overburdened moderation replies while the rumor mill does its work?

Does the AI generated art I post be used as the argument against me?

See, there was the case among the artists (i.e. people who draw pictures) where they falsely accuse other using AI as the way to tarnish someone's reputation. They are much more sensitive than this site is, and we I doubt we need to bring their witch hunts here.

Just go to the Royal Road and look what toxic hellhole in turned into when one allows writers to sabotage their competition for sake of profit.

...Flag it, and the people running the site will review the novel, send it through a AI checker, and if all signs point to yes, they will ask the writer to send in proof that the writing was not AI influenced, or they will add the 'AI-edited' tag.

Sounds reasonable to me?

And it's not like not adding this feature will remove hateful accusatory people. Since this is the case, why not give authors an option to say it's AI edited if it is? And the option to say it's not if it's not, and the mods not tagging it is the proof? Instead of being review bombed?

Lastly, such a transparent environment will attract readers. Authors need readers, lmao.

 

[CountVampyre]

...Flag it, and the people running the site will review the novel, send it through a AI checker, and if all signs point to yes, they will ask the writer to send in proof that the writing was not AI influenced, or they will add the 'AI-edited' tag.

Sounds reasonable to me?

And it's not like not adding this feature will remove hateful accusatory people. Since this is the case, why not give authors an option to say it's AI edited if it is? And the option to say it's not if it's not, and the mods not tagging it is the proof? Instead of being review bombed?

But who will check it? Tony is all alone...

 

[StoneInky]

But who will check it? Tony is all alone...

I didn't know that? Story covers and new stories get approved and flagged on the site all the time, so I assumed we had more staff.

...But I still stand my ground. Even if you can't enforce it, Tony, just add it as another content warning tag/mark.

 

[beast_regards]

...Flag it, and the people running the site will review the novel, send it through a AI checker, and if all signs point to yes, they will ask the writer to send in proof that the writing was not AI influenced, or they will add the 'AI-edited' tag.

Sounds reasonable to me?

And it's not like not having this feature will remove hateful accusatory people. Since this is the case, why not give authors an option to say it's AI edited if it is? And the option to say it's not, and the mods not tagging it is the proof? Instead of being review bombed?

It doesn't seem feasible.

The Scribble Hub, as it is, has only one moderator, Tony, its owner, who at this point needs not only to reply to the various reports sent to him but also needs manually approval each cover, each review, each chapter (1st chapter, at minimum).

Now, he would have to deal with the myriads of accusation of "using AI", while we, as the users, could barely agree on what "AI" even means. If you take it to the corporate marketing extreme, even the Windows 11 is "AI" not to mention the various spellcheckers built into Google Chrome, or other browsers, as well as the programs like Pro-writing Aid, or Grammarly, and many, many more.

In fact, this post was made using AI?

How, well....

Pro-Wiriting Aid tells me right now this post sucks. It has only 75% Spelling Rating

Additionally, not everyone uses Google Docs even. Some even have the accursed Word installed in their machines, and if the Microsoft marketing told me anything, it's AI too, even before you write anything into it.

 

[Phantonym]

This reminds me of a quote: "I want AI to do my laundry and dishes so that I can do art and writing, not for AI to do my art and writing so that I can do my laundry and dishes."

 

[beast_regards]

...But I still stand my ground. Even if you can't enforce it, Tony, just add it as another content warning tag/mark.

People don't tag their stories properly.

Hell, my own stories have reviews complaining I don't label my story properly, imagine the review bombing that would follow should anyone be convinced that anything is "AI" ....

There has been even a motion to delete the fan-fiction from the site, and well, guess what, no one labels that correctly either, and it still doesn't stop people from demanding deletion.

Now imagine proving your ideas are in fact you're in addition to making sure you aren't using forbidden "AI" software. Remote access to someone's machine in addition to forcing to reveal their identity, that is starting to stink...

Not to mention, even if everything it is "just a tag" ...

No one would use the tag that would damage his story (or perception of his story).

On the Royal Road, it's even for the best no not use any content warning at all, as the tag enabled makes your story invisible without the age verification, and no one wants to be invisible. And their warning range from "violence" to "sexual" to "AI generated" to "AI assisted" ...

It's a witch season out there.

 

[3guanoff]

TL;DR: AI warning? aye! suspicious drama; tags, etc. some misuse but useful: better content discovery.

Bit dramatic up there. The proposal is for adding a content warning. Of course, some people might use it incorrectly. People tag their story incorrectly. Yet we have not removed tags, genres, and content warnings.

While not everyone uses such features correctly, enough people do that they are still useful. An AI warning would indubitably be useful. It gives honest people an opportunity to avoid bad ratings by disappointed readers who dislike AI. It also gives writers an opportunity to attract readers who do enjoy certain aspects of such contents: i.e., fast releases, better(?) adherence to grammar and style, and the ostensibe lack of typos.

 

[JayMark]

It's going to become nebulous drawing the line eventually.

Only once AI is indistinguishable from human writers. At which time I will never pick up a pen creatively ever again.

Is using the magic wand in photoshop AI? Why isn't it?

No, it's photoshop and it predates AI. That's like saying video editing is AI, it's not.

Phone predictive texts? Because that's AI.

Who would use phone predictive text for writing? My phone hinders what I'm trying to say more often than it helps.

Spell checkers? AI. Professional editing software? AI.

Spell checkers work off a database. They existed before AI and are not AI. Grammar checker is just a database program that predates AI as well, it is not AI. Also, the grammar checker on microsoft word forces a conformity and even blatently wrong grammar on me by making me look at the stupid blue squiggly line.

Name generators? AI too.

Nobody is against using AI for research, especially is it is checked against more reliable sources. That said, since AI halucinates it is not reliable for name generation without researching and fact checking.

So many writers use AI covers without a care, but get angry at AI writing, and vice-versa.

Yes, this is a hypocrisy. AI 'art' keeps getting better too. I was excited to use it at first but now I feel less enthusiastic about it.

Honestly, most work on this site should have a "AI assisted me in some way."

At which point it becomes meaningless.

I was writing on typewriters before AI existed. Are typewriters AI?
But the question isn't about completely eliminiting AI, it's about those who have entire texts, chapters, paragraphs generated by AI and try to claim they've written something. I use an AI cover atm, but if I go around saying 'look what I made/painted/drew' I expect to take abuse for it because that's lying.

 

[Leti]

No, it's photoshop and it predates AI. That's like saying video editing is AI, it's not.

That's like saying the original man of steel use AI. He isn't.

 

[beast_regards]

I was writing on typewriters before AI existed. Are typewriters AI?

By today's logic? Yes, they are.
My coffee maker is "AI" too.
Which makes this effort entirely pointless. As I said so many times, the word, or rather abbreviation, was used so many times, often out of context, that it lost its original meaning and become nothing more than the buzzword you could conveniently throw at your opponents.

That's like saying the original man of steel use AI. He isn't.

Remove the man. Remove the problem.

TL;DR: AI warning?

AI warning is pointless.

The absence of the common definition aside, the site, as it is, has this covered through their content rules, which states that any novels purely generated by the AI are forbidden on site in the first place, and would be rejected by Tony immediately after being submitted. See, the approval process is completely manual, and thus any story that passes the metaphorical Tony's sieve should be considered admissible as far as the AI content (and US law) goes.

Adding the tag as this point would put the approval process in question, and would most likely be used as the way to actively harass the competition rather than filter out any harmful content.