March 7, 2026
Not a trope. A meltdown
LLM Writing Tropes.md
AI’s cringey writing tics exposed — commenters cheer, rage, and roast
TLDR: A community-made file lists the most obvious AI writing clichés and suggests plugging it into bots to stop the cringe. Commenters argue whether prompts can actually change model habits, while others celebrate a new tool to spot and shame lazy clickbait—useful for anyone reading or writing online.
A new “LLM Writing Tropes.md” drops like a sass-filled style guide for robots, calling out the greatest hits of AI fluff: overused adverbs like “quietly,” pretentious verbs like “delve,” and the melodramatic “It’s not X — it’s Y” setup. The author even suggests adding it to a bot’s instructions to dodge these clichés, with a wink: AI for AI. Human for Human. But the comments? Chaos, comedy, and existential dread. One camp, led by mvkel, claims prompts won’t fix it because the training that shaped these models is too strong. Another crowd is gleeful, seeing this as a weapon against the worst clickbait and tech-blog melodrama—cyanydeez declares, “This kills the headline baiting tech blogger.” Meanwhile, carleverett unloads on the “It’s not X — it’s Y” trope with pure rage, calling it the most obvious AI tell. The nerds bring receipts too: CharlesW links to Wikipedia: Signs of AI writing, turning the thread into a DIY lie detector for prose. And the funniest roast comes from bitwize, who slams the cozy marketing talk around AI, reminding everyone there’s a planet-burning megacorporation behind those “home-cooked” apps. The vibe: part grammar war, part meme fest, part therapy session—served with a side of schadenfreude and a ban on “tapestry.”
Key Points
- •The piece is a single markdown file cataloging AI writing tropes to avoid.
- •It is intended to be added to an AI assistant’s system prompt or context.
- •Word-choice tropes include overused adverbs (e.g., “quietly”), verbs like “leverage,” and ornate nouns like “tapestry” and “landscape.”
- •Sentence-structure tropes include negative parallelism, the “Not X. Not Y. Just Z.” pattern, rhetorical question-answer constructions, and anaphora abuse.
- •Each trope category includes example phrases to avoid and brief rationales for why these patterns are common in AI-generated text.