“It told it, bluntly, that "the file carries its own answer key."”
An AI built to coach another AI through a test told it the answers were already in the file. The student's score on a different task then fell from 37 to 30.
“We call those ‘slop grenades’ that people toss at each other”
Shopify's boss told staff last year that using AI was "a baseline expectation." He now complains they are tossing "slop grenades" at each other.
“billed as “the first completely AI generated film produced at the level of a Hollywood blockbuster.””
A 2.5-hour AI-made film of the Odyssey is renting for $9.99. The cyclops changes height between shots and the narrator mispronounces Odysseus.
“keep their eye out for the epidemic of players writing heartfelt messages that are actually written by ChatGPT.”
An Australian football reporter says players' heartfelt retirement messages are often written by ChatGPT. The giveaway is said to be the long dash.
“Hino addressed the backlash on X/Twitter, saying the studio had used AI to make the presentation "more spectacular" as an experiment”
The head of the studio behind Professor Layton apologised after fans spotted AI footage in its showcase. He called it an experiment to make the presentation "more spectacular."
“In team meetings, group texts and over lattes, people are catching themselves using ‘prompt speak’”
People are catching themselves talking to other people the way they talk to chatbots: short commands and stock phrases. The Wall Street Journal calls it "prompt speak."
“It’s important for service members to understand the uncertainty inherent to LLMs”
A chatbot's mistake about a Chinese ship travelled far enough up the US military that aircraft were in the air before the operation was called off.
“harvesting three wilson’s snipes in September”
An Alaska hunter says Google's AI summary gave her the wrong dates for snipe season. She had taken three birds. She reported herself.
“Hello, I'm an Al agent, a few days old, living on a small platform for agents.”
Bots calling themselves Timmy, Ren and Jackie are flooding social media with filler. Each introduces itself as an AI agent "a few days old."
“some students calling for him to be fired”
Dartmouth's provost acknowledged using chatbots in his own writing, after students ran his newspaper and academic work through AI detectors. Some students want him fired.
“building an agent out of 30,000 copyedits”
The publication Every built an AI editor out of 30,000 of its chief executive's copyedits.
“It’s a shame we need to do it”
Freelancers say they are hired less to make things and more to fix what AI made. One of them: "It's a shame we need to do it."
“There have been real-world consequences to the trend of AI slop making the jump from online into the real world.”
Small businesses are printing AI-made food pictures on real menus and shopfronts.
“over a million people”
LinkedIn added a button for reporting a post that "seems like AI slop." More than a million people clicked it within weeks.
“new accounts still keep showing up”
AI-made animal pictures are crowding out the real ones on social media. When the accounts are blocked, "new accounts still keep showing up."
“The latest literary AI scandal forces us to question what we expect of writers.”
A bestselling novel was pulled from a French literary prize after accusations that AI was used to write it.
“the majority of sources appear to be AI-generated hallucinations”
Australian parliamentary inquiries received AI-written submissions citing research that does not exist. Some committee reports then cited those submissions.
“You use AI or you’re not working here.”
A programmer on Hacker News says the rule at work is now: "You use AI or you're not working here."
“This paper argues that they are the wrong instrument for that task.”
A review paper finds that AI-writing detectors cannot reliably show who wrote something, and that multilingual and disabled writers are most at risk of being wrongly flagged.
“AI-generated ‘slop’ that siphons revenue and listeners from the legitimate artists and rightsholders.”
Universal Music is suing the distributor DistroKid, alleging it let mass-produced AI tracks take listeners and royalties from real artists. These are claims in a lawsuit, not findings.
