In a story they could have predicted in their own work, a sci-fi magazine has asked people to stop sending stories generated by artificial intelligence in the latest sign of apps like ChatGPT gaining prominence.
Clarkesworld, an award-winning science fiction and fantasy publication from New Jersey, has temporarily suspended all submissions for having too many written by AI.
The irony is not lost on the magazine, which has a robot as its mascot.
Publisher and editor-in-chief Neil Clarke shut everyone down on Feb. 20 after being able to tell that hundreds of stories weren’t written by humans.
“By the time we closed on the 20th, around noon, we had received 700 legitimate submissions and 500 machine-written ones,” Clarke told NPR.
Clarkesworld, an award-winning science fiction and fantasy publication from New Jersey, has temporarily suspended all submissions because too many are written by AI
While he wouldn’t explain how they could tell the stories were made with AI, he generally stated that the writing was clearly bad.
“It increased so quickly that we thought we would have double the number of entries than usual by the end of the month,” he added.
“And given the rate at which it had grown compared to previous months, we were concerned that we had to do something to stop it.”
Clarke notes that his magazine pays its writers, meaning people are trying to make money without actual writing.
“There is a rise of online crowds,” he said. “And some people have followers who say, ‘Hey, you can make some quick money with ChatGPT, and here’s how, and here’s a list of magazines you could sign up for.’ And unfortunately we’re on one of those lists.”
Clarke said the magazine doesn’t yet have an answer on how it’s going to address the issue, and part of the motivation for speaking out was hoping to crowdsource some solutions.
He adds that, like much fantasy work, it offers a worrying glimpse into the future.
Publisher and editor-in-chief Neil Clarke (pictured) cut everyone’s mouth on Feb. 20 after he was able to tell that hundreds of stories were not written by humans
Clarke notes that his magazine pays its writers, meaning people are trying to pocket money without any actual writing
Clarke said the magazine doesn’t yet have an answer on how it would approach the issue, and part of the motivation for speaking out was hoping to crowdsource some solutions.
“But the thing is, science fiction is often a cautionary tale, and you know, we don’t embrace technology just because it exists. We want to make sure we’re using it right.
“And there are some important legal and ethical issues surrounding this technology that we don’t want to accept,” he noted.
Clarkesworld’s Twitter account posted on Wednesday that submissions will likely be allowed again “sometime next month.”
OpenAI invented ChatGPT, a startup co-founded by billionaire and Twitter CEO Elon Musk, and is headquartered in San Francisco.
A system prototype went online on November 30, 2022 and quickly climbed the popularity ladder, with over a million users in its first week.
ChatGPT uses a machine learning technique called Reinforcement Learning from Human Feedback (RLHF).
This means that it learns through interactions with its environment.
The model was trained with 570 GB of data collected from books, web texts, Wikipedia articles, and other online writings.
ChatGPT launched in November 2022 and has been growing in popularity ever since. The AI-powered chatbot generates emails, talks naturally to customers, and provides patients with medical information
While experts think ChatGPT will replace Google within two years, the AI-powered chatbot thinks otherwise
Chatbots have been filtering the internet for decades, but can only answer simple questions with standard answers.
But ChatGPT is considered the future of these systems, conducting conversations with users through multiple questions, providing lightning-fast answers, and generating software code.
Fears are spreading in Silicon Valley that ChatGPT – the AI chatbot taking the world by storm – could become the world’s go-to search engine.
Google execs reportedly declared a “code red” over fears that the $150 billion-a-year search monopoly could be wiped out thanks to the Microsoft-backed tool.
Much has been written about ChatGPT’s ability to perform eerily human professional tasks such as writing emails and resumes. But the fear in big tech stems from the fact that it can instantly respond conversationally to user queries, using data collected from the internet.
That’s a concern for search engines that rely on users scrolling and researching themselves, exposing them to ads. However, experts have noted that ChatGPT seems to have a liberal bias and the answers are not entirely transparent, casting doubt on its effectiveness as a search tool.
