AI-generated fiction is flooding literary magazines—but not fooling anyone

A short story called “The Last Hope” hit Sheila Williams’ desk in early January. Williams, editor Asimov science fiction free, magazine, review the story and forward it.

At first, she didn’t think much of it; she reads and responds to writers daily as part of her job, receiving between 700 and 750 stories per month. But when another story, titled “The Last Hope”, appeared a few weeks later by a writer with a different name, Williams was skeptical. By the time another “The Last Hope” arrived a few days later, Williams knew immediately that she had a problem on her hands.

“That’s like the tip of the iceberg,” says Williams.

Since that first submission, Williams has received more than 20 short stories called “The Last Hope,” each coming from different author names and email addresses. Williams believes they were all generated using artificial intelligence tools, along with hundreds of other similar submissions that have plagued small publishers in recent months.

Asimov it received about 900 stories for consideration in January and is on track to receive 1,000 this month. Williams says that almost all of the increase can be attributed to pieces that appear to have been generated by AI, and she’s read so many that she can now often tell from the first few words whether there’s a chance it’s not something written by someone.

Sometimes they don’t even bother to substitute “(name)” for their own

Aside from repeating titles, there are certain character names that appear frequently, according to Williams. Sometimes the manuscript will contain a different title is the one shown in the online form. Author names often appear to be a combination of first and last name. In optional cover letters, some authors include instructions on how to wire them money for their not-yet-accepted story. Sometimes, the submitter doesn’t bother to replace “(name)” with theirs.

Using ChatGPT, The Championship he was able to replicate some aspects of the submissions seen by Williams. A tip for writing a science fiction short story — complete with copy-and-paste information Asimov submission guidelines — he combined stories with many similar titles in a row, such as “The Last Echo,” “The Last Message,” “The Last Day of Autumn,” and “The Last Voyager.”

Willams and her team learned how to spot AI-generated work, but the submissions were the same. Outlets like Asimov going overboard with AI chum, fighting for the time of editors and readers and possibly crowding out genuine submissions from newer writers. And the problem could only get worse, as the wider availability of writing bots creates a new genre of get-rich-quick schemes, where literary magazines with open submissions have found themselves on the receiving end of a new surface for Spammy submissions asking for a game. the system.

“I basically go through them as fast as I can,” Williams says of the pieces she suspects are AI-generated. “It takes the same amount of time to download a submission, open it and view it. And I’d rather spend that time on the legitimate submissions.”

For some editors, the influx of AI-generated submissions has forced them to stop accepting new work.

Clarke believes the submissions are coming from influencers and “ghost” websites.

Last week, the popular science fiction magazine Clergy world Announced would temporarily close submissions due to a flood of AI-generated work. In an earlier blog post, editor Neil Clarke noted that the magazine had been forced to ban a number of skyrocketing authors because they had submitted stories generated using automated tools. In February alone, Clergy world There were 700 human-written submissions and 500 machine-generated stories received, Clarke says.

Clarke believes the spammy submissions are coming from people looking to make a quick buck and have found Clergy world and other publications through influencers and “ghost” websites. One website, for example, is loaded with SEO bait articles and keywords related to marketing, writing and business and promises to help readers make money fast. An article on the site lists nearly two dozen literary journals and websites – including Clergy world and Asimov, as well as larger outlets such as the BBC with pay rate and submission details. The article encourages readers to use AI tools to help them and includes affiliate marketing links to Jasper, AI authoring software.

Most publications pay small rates per word, around 8 to 10 cents, while others pay flat fees of up to a few hundred dollars for accepted pieces. In his blog, Clarke wrote that there was a “high percentage of fraudulent submissions” coming from some regions but declined to name them, worried that he might paint writers from those countries as scammy.

But the possibility of getting paid is a factor: in some cases, Clarke has corresponded with people banned from submitting AI-generated work, saying they need the money. Another editor told The Championship that they would receive submissions and emails from writers in countries where the cost of living is lower and an $80 publication fee goes far beyond what it goes in the US, even before the stories generated by the AI.

Clarke, who built the submission system used by his magazine, described the efforts of AI story spammers as “selfless” – by comparing notes with other editors, Clarke was able to see that the same work was being submitted from same IP address to multiple publications. minutes apart, often in the order in which magazines are listed.

“If these were people from the inside of the (science fiction and fantasy) community, they would know it wouldn’t work. It would be immediately clear to them that they couldn’t do this and expect it to work,” says Clarke.

The issue extends beyond science fiction and fantasy publications. Flash fiction online it embraces a range of genres, including horror and literary fiction. On February 14, the outlet attached a notice to its submission form: “We are committed to publishing stories written and edited by people. We reserve the right to reject any submission that we suspect is generated or created primarily by language modeling software, ChatGPT, chat bots, or any other AI apps, bots, or software.”

The updated terms were added around that time FFO received more than 30 submissions from one source within a few days, says Anna Yeatts, publisher and founding co-editor. Each story involved cliches that Yeatts had seen in Yeatts-generated work, and each had a unique cover letter, structured and written unlike what the publication usually sees. But Yeatts and colleagues had suspected since January that some work sent to them had been created using AI tools.

Yeatts had been playing around with ChatGPT starting in December, feeding the tool’s tips to produce stories about specific genres or in styles like gothic romance. The system was able to replicate the technical aspects, including establishing main characters and setting up and introducing conflict, but failed to produce any “deep perspective” – ​​the endings were too neat and perfect, and emotion often entered the melodrama. Everyone has “piercing green eyes,” and stories and characters often open up sitting down. Of the more than 1,000 works FFO found this year, Yeatts estimates that AI is likely to have generated about 5 percent.

“We put that scary little warning up (on the submissions page),” says Yeatts. However, enforcing it can be challenging.

In the past, FFO has published mainstream work with a more traditional writing style and voice that is accessible to a range of reading levels. For this reason, Yeatts says that stories generated using AI tools could find past essentials.

“Every part of the story is what you try to look for. It has a beginning, middle, and end. He has a secret, characters. The grammar is good,” says Yeatts. The FFO team is working to train staff readers to look at certain aspects of the story when taking the first pass on submissions.

“We really don’t have good solutions.”

Yeatts is concerned that a growing wave of AI-generated work could put a stop to written work. The outlet uses Submittable, a popular submission service, and FFOa plan that includes a monthly limit on stories, after which the portal closes. If hundreds of people submit ineligible AI-generated work, that could prevent human authors from submitting their stories.

Yeatts isn’t sure what the magazine can do to stop the stories. Upgrading the Receivable plan would be expensive FFOwhich “runs on a shoestring budget,” says Yeatts.

“We’ve talked about soliciting stories from other authors, but then that doesn’t really reflect who we are as a publication because that will deter new writers,” says Yeatts. “We really don’t have good solutions.”

Others in the community are keeping track of the problem affecting other publishers and thinking of ways to respond before it spreads further. Matthew Kressel, a science fiction writer and creator of Moksha, an online submission system used by many publications, says he has started hearing from clients who have received spammy submissions that appear to have been written using AI tools.

Kressel says he wants to keep Moksha “agnostic” about the value of submissions generated using chatbots. Publishers have the ability to add a checkbox where writers can declare that their work doesn’t use AI systems, Kressel says, and he’s considering adding an option for publications that would allow them to block or partially limited using AI tools.

“Allowing authors to assert themselves if the work is generated with AI is a good step,” Kressel said The Corr by email. “It provides more transparency to the whole thing, because there are a lot of uncertainties at the moment.”

For Williams, editor of Asimov, forcing her to use her time to sift through the AI-generated junk is frustrating. But it’s even more worrying that new, legitimate authors see what’s happening and think that editors will never pass their manuscript.

“I don’t want writers to worry that I’m going to lose their work because I’m inundated with junk,” says Williams. The good news is evident very soon. “The mind that cultivates the interesting story is in no danger.”

Leave a Reply

Your email address will not be published. Required fields are marked *