So it begins. Submissions for new authors closed due to AI entries

LA Thomas

Admin help

  • LA Thomas
  • Replies: 6
Peyton Stafford

Reality Check Structuring beta reads, blurbs, and other critiques

  • Peyton Stafford
  • Replies: 6
Status
Not open for further replies.
How depressing, and yet predictable. As if this game isn't a tough enough slog already. :angry-face:

More worrying is the publisher can't immediately identify a fix for filtering out this deluge of dross.
 
Yep. My husband's department has made the decision to go back to paper and pencil exams after investing a ton of time and effort into online exams (that work better for both students and staff) over the past few years, because the exam answers AI can give are better than most of the students can do, and while it's possible to weed out AI, no one has extra time to put in to do so. The fix for magazines isn't so simple--I certainly wouldn't want to be the one trying to navigate this as a publisher.

I'm also thinking I may put on hold my magazine submissions for the time being... why bother when you know they're all dealing with this deluge of AI submissions?
 
@Robinne Weiss , maybe we should go back to hand written stories for submissions? ha! (jokes, kind of.)

The creators of these chat bots have a lot to answer for. I don't buy that they just create stuff and aren't responsible for what's done with it. There has to be a greater level of care and responsibility for tech moving forward. If there's not, we're all sitting ducks. Grrrrrr.
 
There's something to be said by asking for handwritten subs. But in my case, I would never have been writing had it not been for my first work supplied laptop.

My handwriting is utterly atrocious and could not have been decoded even by those who worked at Bletchley Park. Sometimes I read notes I've made on a pad for plot points or shopping lists and am baffled. Like the time I went to get a bag of sugar and came back with a kidnapped racehorse.
 
I saw a post from an agent who supposedly got her first AI written query letter! I think it's everywhere now. Only a matter of time before we get an AI-written Pop-Up Submission!
 
Someone needs to code a piece of software quickly that can spot this crap or we're all going to suffer.

It should be our inalienable right to be the sole purveyors of dross to publishers and agents. Ours and no one or nothing else's.
 
I think there are some lawsuits being pushed through - or at least some people trying to push them through - regarding copyright and AI right now. My husband is a chip designer in the AI field and is actually really excited about the lawsuits. He is really frustrated about how there are no parameters and it's just an open minefield right now. He is seeing it with resumes actually. He gets amazing resumes and then people come in and can't code for sh!t he says, so he knows it's been generated. It's a huge time waster. It should be interesting to see how lawmakers address this.
 
Jonny said:
Someone needs to code a piece of software quickly that can spot this crap or we're all going to suffer.
Click to expand...

Search engine companies are trying to do that. And probably also the AI companies. From what I've read nobody is really quite sure what happens when the AI picks up more and more AI-generated text from the net and uses it to produce new text that will be published on the net which will then be picked up ... etc.
 
JohnBertel said:
Search engine companies are trying to do that. And probably also the AI companies. From what I've read nobody is really quite sure what happens when the AI picks up more and more AI-generated text from the net and uses it to produce new text that will be published on the net which will then be picked up ... etc.
Click to expand...

That’s a really grim scenario.
Like an avalanche of an exponential quantity of AI generated slush gumming up the system.
 
I do not see any reason why someone could not build an AI that would examine submissions for errors that AI bots might make, such as boring and predictable plots or stock characters, and weed them out.

Sales will tell whether AI can compete with humans.
 
Peyton Stafford said:
I do not see any reason why someone could not build an AI that would examine submissions for errors that AI bots might make, such as boring and predictable plots or stock characters, and weed them out.

Sales will tell whether AI can compete with humans.
Click to expand...
My husband says this already exists although they aren’t very accurate. He says it’s also easy to write a program to thwart it. So basically… the safeguard is there, but the AI output is too unpredictable to make it effective and people are too underhanded too make it effective.
 
Having an interest in this field myself, I reckon that they'll never be able to fully detect and weed out AI content technically. There's a couple of likely scenarios. Publishers will get legal, asking authors for an indemnifying affidavit stating that the work is all their own (similar to copyright). In general, a market for authentic content will emerge (the "vinyl" of writing!). Publishers and magazines may also move more towards invite-only submissions.
 
Peyton Stafford said:
The Authors Guild modified their standard contract to include a clause stating the work cannot be used to train AI. That is on the assumption publishers will use their writers' books to train AI to write similar books for free.
Click to expand...
I don’t think books are being fed into any AI stuff. My understanding is that if it’s put online it’s free reign, but I could be wrong on that with new AI tech development.
 
UGH. of course. AI-generated everything. my friends in the art community are having to deal with AI art generation tools like NightCafe being used to mimic their own art styles -- "instead of paying for a commission from an artist, why not just feed their work into an AI and get something of the same style for free?"
what makes these things extra terrible is the fact that there everyone and their dog can hop on an AI generation site and crank out fifty of whatever they want in minutes. chatgpt, character.ai, chai, etc. can be used for whatever you want. there's even sites made specifically for generating writing -- see rytr.
the thing is, these sites weren't created maliciously. none of them are flashing neon signs saying "CLICK HERE TO HAVE YOUR ESSAY WRITTEN FOR YOU!" or "USE THIS EASY SITE TO COPY OTHER ARTISTS!" all of these sites were made for creative or entertainment purposes -- a lot of AI art generation sites are made just to show people "look! robot made picture! so cool!" and nothing else. rytr is for copywriting -- you enter a paragraph or a cover letter or an email that you wrote, and it modifies/checks/rewrites it to be better. people are just taking advantage of these sites, and they always will until they're taken down, which really sucks for people like me who check out these sites just to mess around with prompts and stuff.
not surprised, just disappointed.
 
what really bothers me about this is
1. people who create things by themselves will start to be accused of using AI, and getting rejected/berated for something they didn't do
2. people who actually use AI will circumvent any and all attempts to catch AI-generated stuff, and they'll share these solutions so quickly and widespread nobody will be able to do anything about it. happens with everything ever -- schools block websites, kids figure out the admin password to get around it, and share it with all their classmates. a chat site bans certain words, users get around the filter through symbols and weird fonts, and the site is 0% cleaner.
i'm really not sure how we're going to fix this one, guys.
 
Brooke said:
what really bothers me about this is
1. people who create things by themselves will start to be accused of using AI, and getting rejected/berated for something they didn't do
2. people who actually use AI will circumvent any and all attempts to catch AI-generated stuff, and they'll share these solutions so quickly and widespread nobody will be able to do anything about it. happens with everything ever -- schools block websites, kids figure out the admin password to get around it, and share it with all their classmates. a chat site bans certain words, users get around the filter through symbols and weird fonts, and the site is 0% cleaner.
i'm really not sure how we're going to fix this one, guys.
Click to expand...
OMG you are so correct with #2! Roblox anyone? Sorry but I had to laugh at that.
 
ChantalS said:
OMG you are so correct with #2! Roblox anyone? Sorry but I had to laugh at that.
Click to expand...
exactly what i was thinking of when i wrote that!! popular game gets banned, and there's 5 more copies of it up the next day
 
I've had voice actors audition for my Audibles seem like their submissions were generated by some kind of voice software, and now this. Sigh.

Well, as I like to sign off, keep writing, and Live the Dream:^)
 
Status
Not open for further replies.
LA Thomas

Admin help

  • LA Thomas
  • Replies: 6
Peyton Stafford

Reality Check Structuring beta reads, blurbs, and other critiques

  • Peyton Stafford
  • Replies: 6
Back
Back
Top