The award-winning Clarkesworld Magazine has supported the careers of science fiction writers for nearly 20 years and regularly features works by Hugo Award nominees and winners such as Elizabeth Bear, Peter Watts and Catherynne M. Valente. But right now, in a rather ironic situation, it’s struggling with the sci-fi among the modern trends: AI.
After to a current article by Clarkesworld Editor Neil Clarke, Over a third of the submissions received by the magazine this year were written by artificial intelligence and then submitted by fraudulent humans. And it’s getting worse, fast. The first half of February saw more than twice as many AI-written entries as all of January, Clarke recounts my city Today alone there were 50.
Since the article was written, Clarke has tweeted that submissions are now fully closed. “I shouldn’t be hard to guess why,” he adds.
The decision to close the filings was made “in a short space of time,” Clarke said my city via email when the numbers came in this morning. “I could either play Whack-a-Mole all day or close submissions and work with the legitimate submissions.”
The rate of rise of this situation is quite striking. Clarke states in his blog post that he’s had to deal with plagiarism for a long time, but it wasn’t until late 2022 that the problem became so endemic. And then, in the first month and a half of 2023, it escalates to such an extent that the magazine suspended entries entirely.
How can Clarkesworld tell that a story was generated by AI?
Clarke doesn’t explain on his blog how he can tell which entries were written by AI, for the very reasonable reason that he doesn’t want to arm cheats with information that might help them bypass his detection. However, he explained my city that they are currently not too difficult to spot.
“The ‘authors’ that we banned,” Clarke said, “obviously submitted machine-generated text. These works are formulaic and of poor quality.” However, he also suspects that there is already a level above that, not quite as obvious, but enough to raise suspicion. “None of these are ever good enough to justify spending more time on them,” he explains, but adds, “It’s inevitable that this group will grow over time and become another problem.”
It’s not a problem Clarke faces alone. The publisher reports that others in similar positions are facing the same challenges, and if it happens to Clarkesworld it will clearly happen anywhere it is open to submissions for publication. And while for the most part such submissions are simply weeded out because they aren’t good enough for publication, sifting through the fakes is an expensive and time-consuming process.
Clarke adds that given the number of false positives and negatives and the cost of such services, third-party detection tools that claim to be able to detect plagiarized or AI-written content are not the answer. Other short-term measures like regional lockdowns on parts of the world where most of the fake listings come from aren’t the answer either. As Clarke puts it in his article,
It is clear that business as usual will not be sustainable and I fear that going this route will result in an increased number of barriers for new and international authors. short story needs these people.
And of course, this is not a problem that will become easier. The pace at which AI chat bots are improving is enough to write ideas for a sci-fi short story, and presumably forthcoming tweaks will make them increasingly difficult to spot right away. However, it’s likely that we’re still a long way from AI being able to write truly readable stories. I asked Clarke if he thought that was likely. “At the moment there is still a clear need for improvement,” he said, without wanting to venture a guess as to how long such a leap could take.
But that’s not much consolation. “We still have ethical concerns about the means by which these works are created,” Clarke said my city“and until such concerns can be resolved, we will not even consider publishing machine-generated works.”
ChatGPT and Chatsonics attempts at a science fiction story
There are already services like ChatSonic boldly touting as a means of creating blocks of non-plagiarized fonts for students to use. I’ve had exhaustive futile debates with the AI myself before that this is clearly a scam, to which it gets enormously outraged, defending itself with circular reasoning and noting that simply asking the bot for words on an issue itself is a creative act .
As I was writing the previous paragraph, I asked ChatSonic to write me a 1,000-word short story about an AI who writes science fiction and later wins a Hugo Award. For some reason it only hit 293 words (damn freelancers), and it sucks, but it took a few seconds:
Meanwhile, ChatGPT put in a lot more effort, hit the word count and wrote something that had some level of creativity behind it. Ultimately, it’s still a horrible story and hilariously self-aggrandizing but unnervingly competent:
(Um, I think I’ll put the second half in the comments if you’re dying to know how it ends.)
Can AI trump human creativity?
Clarke mentioned above that he has many ethical concerns to resolve before even considering publishing AI-made writing. But could something like this ever happen? If AI could generate original stories worth reading, would it ever make sense to publish such things? “First of all,” Clarke told us, “you need these tools to be able to write something beyond the data set. Real fantasy, not a remix. At this point it can competes with our best writers, but isn’t necessarily better.”
Of course, “better” might not be the deciding factor. Clarke adds: “The big difference that’s giving us problems now is speed. A machine can surpass and bury a human artist in the noise of it all.”
And just in case all of this hasn’t worried you enough, let’s finish things off with ChatGPT’s chilling closing paragraph on the short story I requested earlier:
Of course, some people were still skeptical. They believed that an AI could never be truly creative, that it would only choke out information that had been programmed into it. But SciFiGenius fans knew better. They knew the AI was capable of so much more than just spitting out canned stories. They knew they were dealing with a true artist, capable of creating works that touched the hearts and minds of millions of people.
By the way, you can Support Clarkesworld Magazine in many ways. This is something that will become even more important when Amazon will discontinue its Kindle subscription services later this year.