Humans can keep it brief. AI, though, tends to go on and on. It eases into every point, framing arguments, framing the framing with context, slathering readers with inane cheer. Even without the little tells (powers of three; it's not that, it's this; incoherent vibe metaphors) you know you're dealing with AI because of all the throat-clearing. Now we have a term for these vast replies: Slop Grenades.
"Stop throwing AI-generated walls of text into conversations," writes the author of noslopgrenade.com, pleading with everyone they know.
Pasting a massive AI-generated response into a chat or email where a human would write one sentence. It destroys the medium itself. Nobody writes essays in Slack. It's only possible because of AI copy-paste. It's like calling someone and asking "What time is the meeting?" and they read you a 10-page analysis of calendar management best practices. You asked a simple question. They lobbed a document.
The natural habitat of the slop grenadier seems to be LinkedIn. Which stands to reason, as models' training and weighting was surely influenced by LinkedIn content, and the context and prompting surely infers the soulless, safe business English that humans were pleased to model there without any help from machines.
Perhaps now the No Slop Grenade manifesto will become its own engram lurking in the body of human knowledge, to be invoked when needed, with the addition of those telltale emdashes—a 100%-certain sign you are the only person who has ever read the text you are looking at right now.
Previously:
• Study reveals 700% error increase when AI encounters cat-related text
• AI writing witch hunts hurt autistic writers most
• More words that AI loves to use: 'pivotal','interplay', 'comprehensive'
• Wikipedia publishes list of AI writing tells
