What if the admin(s) were to create a forum entitled something like "Potentially authored by AI?"

[AG31]

What if the admin(s) were to create a forum entitled something like "Potentially authored by AI?"

A flagged story would be posted there by the admins (only admins could start a thread) and we could all register our opinion as to whether it sounded AIish. This wouldn't be a decision making forum, but it would provide information for @Laurel, or whomever, to review whatever process labeled the particular story as AI.

What do you think?

 

[Bazzle]

I would say no...
Because so many *bright sparks* would play games.
It's bad enough with one bombs, but putting all the stories of dubious intent on one place would mean some peeps miss out.

Rather than having "my AI generated story has been rejected for AI"
We would have "my AI generated unloving wives has been put in with the romance AI?" Thing.

Try and keep the door shut for as long as possible. If you open it an inch it will take a mile

 

[AG31]

Rather than having "my AI generated story has been rejected for AI"
We would have "my AI generated unloving wives has been put in with the romance AI?" Thing.

My thought was that only the admins would be able to start threads in this forum.
The replies wouldn't be votes. They would be reasons why they did or didn't agree that AI was being used. The admins would take it as information.

 

[TxRad]

Admin, which consists of one person, doesn't have time for such.

 

[AG31]

Admin, which consists of one person, doesn't have time for such.

Well, it could be seen as a method for reducing the time they spend on identifying AI authorship.

 

[AwkwardlySet]

Laurel doesn't really care for our opinion about the authenticity of the stories, or anything else, really . Your idea, as helpful as it seems on the surface, would just serve as a place where people would express their displeasure at being rejected and then hope to pressure her with the opinions of other users. There is nothing she would gain by doing this.

 

[Bramblethorn]

Well, it could be seen as a method for reducing the time they spend on identifying AI authorship.

I'm not sure it would, though.

Let's say Laurel posts a story in this thread for that purpose. I think a likely outcome is that about a dozen people would express their opinions, we'd argue back and forth for maybe fifty posts about whether it's AI-generated, and not reach consensus, because when do we ever reach consensus on stuff like this?

Laurel then has to read through fifty posts of forum discussion - quite possibly longer than the original story - trying to figure out what the prevailing opinion is, and how much of that opinion is genuine. (If an author is trying to slip an AI story through, they're probably not above creating a few alt accounts to support their own story.)

We've already had a fair bit of bitterness between posters here disagreeing about stories even when nothing actually rides on that argument. If it actually had some influence over the outcome, I think it'd likely generate even more bad feeling.

 

[XerXesXu]

Since we don't know what AI detectors detect, all we would express is our 'feels' about what ought to be described as AI-created rubbish. You can't argue an 'ought' into an 'is'.

 

[Bramblethorn]

Since we don't know what AI detectors detect, all we would express is our 'feels' about what ought to be described as AI-created rubbish. You can't argue an 'ought' into an 'is'.

Story I once heard, undoubtedly apocryphal: in ancient Egypt, the length of a cubit was based on the Pharaoh's forearm. But most people weren't allowed to see the Pharaoh, let alone measure his arm.

So instead, they asked a bunch of people (who'd never seen the Pharaoh) how long they thought his forearm was, and then took the average of those guesses.

 

[AG31]

I'm not sure it would, though.

Let's say Laurel posts a story in this thread for that purpose. I think a likely outcome is that about a dozen people would express their opinions, we'd argue back and forth for maybe fifty posts about whether it's AI-generated, and not reach consensus, because when do we ever reach consensus on stuff like this?

Laurel then has to read through fifty posts of forum discussion - quite possibly longer than the original story - trying to figure out what the prevailing opinion is, and how much of that opinion is genuine. (If an author is trying to slip an AI story through, they're probably not above creating a few alt accounts to support their own story.)

We've already had a fair bit of bitterness between posters here disagreeing about stories even when nothing actually rides on that argument. If it actually had some influence over the outcome, I think it'd likely generate even more bad feeling.

Yeah. I've given up on this idea.
My position now is that we should stop trying to identify and exclude AI generated fiction. Instead we should have a category for it and trust that most authors will be honest. The day will come when people are acknowledged for their talent and creating prompts for AI.

 

[CalebZhass]

I'm not sure it would, though.

Let's say Laurel posts a story in this thread for that purpose. I think a likely outcome is that about a dozen people would express their opinions, we'd argue back and forth for maybe fifty posts about whether it's AI-generated, and not reach consensus, because when do we ever reach consensus on stuff like this?

Laurel then has to read through fifty posts of forum discussion - quite possibly longer than the original story - trying to figure out what the prevailing opinion is, and how much of that opinion is genuine. (If an author is trying to slip an AI story through, they're probably not above creating a few alt accounts to support their own story.)

We've already had a fair bit of bitterness between posters here disagreeing about stories even when nothing actually rides on that argument. If it actually had some influence over the outcome, I think it'd likely generate even more bad feeling.

It would be as simple as using either or on the initial post as a vote, so no, tallying opinions would not be too difficult and anyone could freely change their minds during the discussion if they are convinced.

I don't necessarily find this a bad idea (in an idealized world), but I am also not sure it would not just devolve into a witch hunt, with people falsely accused based on their style and desperately trying to prove that, no.. it's not the AI.. it is them being "this bad".

That is inhumane and degrading. Over the past few days I've learned to accept the current stance of the site and that it will likely stay like this until the paranoia dissipates. You cannot differentiate AI from bad or stale writing and frankly, it should not be anyone's privilege but the reader's to decide what is bad and what is good.

I mean, the ultimate privilege is the site owner's, as they can just refuse to publish something for whatever reason they want, I simply think that overall, in this stage of the site's life, holding back stories hurts it more than just publishing them and letting the audience decide / taking down those who get complaints.

 

[ElectricBlue]

Yeah. I've given up on this idea.
My position now is that we should stop trying to identify and exclude AI generated fiction. Instead we should have a category for it and trust that most authors will be honest. The day will come when people are acknowledged for their talent and creating prompts for AI.

If people are going to be so genius at writing prompts, why don't they just write the bloody thing themselves?

There might be a place for AI prompt speak in business writing (but not in technical writing, because AI makes shit up), but all it would do in fiction is end up in an endless regurgitation of bland, repetitive junk. Where's the creativity and originality in that? It would only encourage mediocrity, and we've got enough of that already.

 

[CalebZhass]

If people are going to be so genius at writing prompts, why don't they just write the bloody thing themselves?

There might be a place for AI prompt speak in business writing (but not in technical writing, because AI makes shit up), but all it would do in fiction is end up in an endless regurgitation of bland, repetitive junk. Where's the creativity and originality in that? It would only encourage mediocrity, and we've got enough of that already.

I honestly do not know. I can tell you from experience, it is more work to tell the AI to write something specific, than it is to write it out yourself. I think there are 4 categories of "AI" claims right now:

The TRUE false positives. People who are simply new or have a very "common" style that is flagged as likely AI for not being special enough.

The tool users. People who use writing aids that - especially if set incorrectly - will mercilessly remove all flair from your writing. Like, just in the previous sentence, word would already want to remove "mercilessly" to "simplify" the text. They end up being false positives again, as even though they wrote their text themselves, the tool had it dumbed down until it again is flagged as likely "AI".

The lazy collaborators. People who use AI to "brainstorm" a story, create a draft, but then don't spice it up sufficiently, so it retains too much of the AI smell.

The cheats. People who just want .. frankly, I have no idea what they want, by uploading to a free story site.. it's not like you are getting paid for your work here. Anyway, people are weird and some would cheat just so they get the recognition for doing something they didn't really do.

---

The last category is the one I find worrisome, but I suspect that their numbers are ridiculously low.
We might have a few lazy collaborators, but I think that's still not too common as it requires very specific tools to pull off.
The other two are probably the vast majority of all "offenders" and as far as I'm concerned, they are all innocent.

Not that it matters what I believe

 

[dmallord]

How about putting some trust in those qualifiers mentioned by Laurel and Manu. 'I affirm this my own creative story, written without artificial assistance. I hold that I have the sole copyright to this submission.'

The statement from Laurel says that the author could face repercussions if it is found to be false. Would that also fall on the Lit organization or could Laurel hold that the onus is on the author and not the site to defend the claim?

This was the process before the AI mess started. Why not continue the 'tradition' of honoring the author's statement of it being 'my creative content and copyright?'

 

[ElectricBlue]

The TRUE false positives. People who are simply new or have a very "common" style that is flagged as likely AI for not being special enough.

The tool users. People who use writing aids that - especially if set incorrectly - will mercilessly remove all flair from your writing. Like, just in the previous sentence, word would already want to remove "mercilessly" to "simplify" the text. They end up being false positives again, as even though they wrote their text themselves, the tool had it dumbed down until it again is flagged as likely "AI".

The lazy collaborators. People who use AI to "brainstorm" a story, create a draft, but then don't spice it up sufficiently, so it retains too much of the AI smell.

The cheats. People who just want .. frankly, I have no idea what they want, by uploading to a free story site.. it's not like you are getting paid for your work here. Anyway, people are weird and some would cheat just so they get the recognition for doing something they didn't really do.

I reckon that's a pretty good summation. I have a lot of sympathy for folk in the first two categories, who are trying to make an honest go at writing, using tools, in good faith and to the best of their ability, to present something reasonable. Their problem, mostly, appears to be a fairly pedestrian style in the first place.

The latter two groupings are the core of the issue, I think - they're basically the same thing: lazy and unimaginative. Back in the day, they'd be hacks, but at least a hack used a typewriter.

 

[astuffedshirt_perv]

I think there are 4 categories of "AI" claims right now:

The TRUE false positives.

The tool users.

The lazy collaborators.

The cheats.

2 more?:
The cover band: takes a well-liked story and 'covers' it. Plot is already built from the original, change names, locations, etc. They get more of what they want to read.

The closer: personally I enjoy writing how protagonist and antagonist ended up making out on the way back to a bedroom, but hate actually writing what happens in the bedroom. I would love to let an AI finish my stories

 

[CalebZhass]

I reckon that's a pretty good summation. I have a lot of sympathy for folk in the first two categories, who are trying to make an honest go at writing, using tools, in good faith and to the best of their ability, to present something reasonable. Their problem, mostly, appears to be a fairly pedestrian style in the first place.

The latter two groupings are the core of the issue, I think - they're basically the same thing: lazy and unimaginative. Back in the day, they'd be hacks, but at least a hack used a typewriter.

I could argue for the collaboration thing, but I already did, so I will rather not repeat it Suffice to say, I agree with you that the lazy part is a problem, even if I do not see the collaboration part a problem myself.

In my book, it's the amount of work and thought they put into it, that differentiates a cheat from a victim.

 

[SimonDoom]

Yeah. I've given up on this idea.
My position now is that we should stop trying to identify and exclude AI generated fiction. Instead we should have a category for it and trust that most authors will be honest. The day will come when people are acknowledged for their talent and creating prompts for AI.

Maybe, but I don't welcome that day, and I have no personal interest in recognizing them for that "talent," which I don't recognize as a talent. I write my stories without help, except things like Grammarly after the fact, and even then I accept Grammarly recommendations cautiously, if at all. My stories do not get rejected for technical errors or AI use because I take the time and effort to avoid those things and to adopt a personal, real-human style. Many other authors here can say the same thing.

This, to me, is another one of those non-problems that don't need a solution. The solution is for authors who are having trouble getting their stories accepted to take more time and effort to make sure they follow the rules and don't get rejected. Some of us, maybe many of us, have absolutely no trouble doing this, and so we're not inclined to be sympathetic to those who do. The answer is not for the Site to go to more time and effort to make things easier for those authors whose writing style reads suspiciously like it's being generated by a computer program.

Any system that is based on "trusting people to be honest," moreover, is, IMHO, no system at all.

 

[CalebZhass]

2 more?:
The cover band: takes a well-liked story and 'covers' it. Plot is already built from the original, change names, locations, etc. They get more of what they want to read.

The closer: personally I enjoy writing how protagonist and antagonist ended up making out on the way back to a bedroom, but hate actually writing what happens in the bedroom. I would love to let an AI finish my stories

The cover, I am not sure it differs form the cheat really. It's still taking the AI's work, trying to pass it off as their own.

The closer is closer P) is either a lazy collaborator or a cheat, depending on how they approach it.

Just to make it clear, you can role play with the AI, assuming the role of a character and let the AI play the other roles, adjusting the script as the story unfolds. That's what I call collaborative writing, where you use the AI to simulate the actions of other characters in the scene until you get a play that you are happy with. You can very much do that with an existing story. Actually, you can mostly only do that with an existing story (or at least the lore of it), as it requires well established characters, plot, and scenes, all of which would have to be created first. Once it is done however, it can be entertaining and gives a whole new twist to writing. Its like using a dice to decide where the enemies will be or what will happen to you next, but a very context sensitive dice, that can react to what you do.

Now the output of something like this is just that.. a series of actions done by the various actors, maybe some dialog. You could try to pass it off as a crude story, but it would stink to hell. I can give you examples if you want, have a few

If you want to turn it into something worth reading, you have to flesh it out. Write proper dialog, scene descriptions, flash out the action, make it consistent in mood, tone, tense, perspective, etc. In short, write your story, as what you have is at best a draft, a storyboard of what happens when.

 

[astuffedshirt_perv]

The cover, I am not sure it differs form the cheat really. It's still taking the AI's work, trying to pass it off as their own.

Plenty of cover bands out there who do not try and pass the music off as their own, and people are more than happy to listen to them.

 

[CalebZhass]

Maybe, but I don't welcome that day, and I have no personal interest in recognizing them for that "talent," which I don't recognize as a talent. I write my stories without help, except things like Grammarly after the fact, and even then I accept Grammarly recommendations cautiously, if at all. My stories do not get rejected for technical errors or AI use because I take the time and effort to avoid those things and to adopt a personal, real-human style.

No, actually, you do not. The AI were trained on text written by real humans and are much better at that than you
You are writing unlike the common folk, which is cool and all, but saying its "real human" completely ignores the whole irony of the situation, where we built machines to mimic us, then other machines to detect if machines are used instead of our own effort and then we now complain that our own effort is detected as something done by a machine.

I see why you feel the way you do, it is a feeling of inequality, I get it. It is fine, I don't want to change your mind. Just pointing out that your style is likely not "real human" if its not caught by the detectors At least not "common real human"

 

[CalebZhass]

Plenty of cover bands out there who do not try and pass the music off as their own, and people are more than happy to listen to them.

I was talking about the types of users that get flagged. You will not get flagged for writing a story based on another story, that's not how this works You get flagged if you write in a particular style. So I took your cover and closer terms and applied them to the topic you replied to. I guess we didn't quite understand each other there.

 

[jhealy55]

Does anyone honestly know what the Anti-AI Rejection Crowd is requesting and why? Even though they get lumped together as being one side of the argument, it feels like it's a pretty mixed bag from where I stand.

My sense is that someone who asks,"show me all the criteria/tools used to evaluate and reject AI." is right up there with someone in the Tour De France asking for a complete breakdown of which substances are tested for and what the threshold value is for being DQ'd. I think the answer is, "don't use any."

It is honestly hard for me to tell which % are truly blindsided by the accusation vs. those who are mad at getting caught vs. those that use a variety of 'tools' but don't think it qualifies as "AI". Given those that are pushing writing aids to college students and those taking their first jobs out of school, I think there may be many who feel like it's widely acceptable in academic and commercial settings, why not at a creative writing site?

I alos think that part of what hampers a lot of people from being sympathetic is that while we all genuinely feel for anyone who is being unfairly rejected, looking at a few of the arguments in the threads, the positions tend to vary along a spectrum of:

- I don't like AI. I don't use AI. I don't want it on the site. Just trying to get my submissions through.
- I don't like AI. I don't use AI. I don't care if others use it. Just want to get rid of the filters in order to get my submissions through.
- I use a little AI to help with editing. I just want a looser setting on the filter so my stuff goes through, even if those using higher AI content get blocked
- Only dinosaurs are not using AI. I don't use it very much, but have no objections if others do. Let's get with the times and stop pretending we're Amish.
- I don't enjoy the writing process without AI, so just quit blocking it. Readers like my stuff even if it's heavily machine produced/edited. Live with it and stop rejecting my stuff!....BRO!
- ANARCHYYYYYYYYYYYYY!

I'm very sympathetic to the first position. Understand the second, but think it's a little selfish to ask to open the floodgates for any one person (we're not executing people so the one innocent man argument doesn't wash for me). I have little to no sympathy for the third position and I'm trending toward hostility the further we get down that list.

 

[CalebZhass]

Does anyone honestly know what the Anti-AI Rejection Crowd is requesting and why? Even though they get lumped together as being one side of the argument, it feels like it's a pretty mixed bag from where I stand.

My sense is that someone who asks,"show me all the criteria/tools used to evaluate and reject AI." is right up there with someone in the Tour De France asking for a complete breakdown of which substances are tested for and what the threshold value is for being DQ'd. I think the answer is, "don't use any."

It is honestly hard for me to tell which % are truly blindsided by the accusation vs. those who are mad at getting caught vs. those that use a variety of 'tools' but don't think it qualifies as "AI". Given those that are pushing writing aids to college students and those taking their first jobs out of school, I think there may be many who feel like it's widely acceptable in academic and commercial settings, why not at a creative writing site?

I alos think that part of what hampers a lot of people from being sympathetic is that while we all genuinely feel for anyone who is being unfairly rejected, looking at a few of the arguments in the threads, the positions tend to vary along a spectrum of:

- I don't like AI. I don't use AI. I don't want it on the site. Just trying to get my submissions through.
- I don't like AI. I don't use AI. I don't care if others use it. Just want to get rid of the filters in order to get my submissions through.
- I use a little AI to help with editing. I just want a looser setting on the filter so my stuff goes through, even if those using higher AI content get blocked
- Only dinosaurs are not using AI. I don't use it very much, but have no objections if others do. Let's get with the times and stop pretending we're Amish.
- I don't enjoy the writing process without AI, so just quit blocking it. Readers like my stuff even if it's heavily machine produced/edited. Live with it and stop rejecting my stuff!....BRO!
- ANARCHYYYYYYYYYYYYY!

I'm very sympathetic to the first position. Understand the second, but think it's a little selfish to ask to open the floodgates for any one person (we're not executing people so the one innocent man argument doesn't wash for me). I have little to no sympathy for the third position and I'm trending toward hostility the further we get down that list.

I simply applied Occam's razor.
- there is no monetary gain from posting a story on the site, so why bother unless you genuinely want to tell something, I see no reason for random people to just spam AI stories, because it is >0 effort for =0 reward.
- using AI to generate decent stuff is harder than people realize, so a lot less people are doing it than the uninitiated would think.
- the way detection works, it simply flags mundane language. Most people use mundane language, hence the many reports of people being flagged.

There is no need to fabricate a conspiracy, when adding all the things together more than adequately explains the situation.

 

[jhealy55]

I simply applied Occam's razor.
- there is no monetary gain from posting a story on the site, so why bother unless you genuinely want to tell something, I see no reason for random people to just spam AI stories, because it is >0 effort for =0 reward.
- using AI to generate decent stuff is harder than people realize, so a lot less people are doing it than the uninitiated would think.
- the way detection works, it simply flags mundane language. Most people use mundane language, hence the many reports of people being flagged.

There is no need to fabricate a conspiracy, when adding all the things together more than adequately explains the situation.

if there's no gain/benefit from posting here, why aren't those getting rejected by moving on and posting elsewhere?