Researchers call out 'LLMish' prose in an important paper
rgblong · x · 2026-10-01
Researcher rgblong voiced disappointment that an important project's paper contains passages in a dense but slippery 'LLMish' register. While not against using AI for writing, he says the style makes it hard to pin down what is actually being claimed, sparking discussion about AI-assisted academic prose.
Related event: Researchers Flag Fabricated Citations and LLM-Style Writing in AI Papers(2 posts)→
More from Fun
- User creates an epic fight scene with sound and VFX using Opus 5.5 — Storge2 · 2026-10-01
- "Some storms learn to fly": a cryptic AI video teaser shared on X — azed_ai · 2026-10-01
- OpenAI's Dot agent flags a wrongly booked Fiji stay, saving user from costly mistake — CtrlAltDwayne · 2026-10-01
- "Open source is for millionaires": dev sparks debate on unpaid coding and class — threepointone · 2026-10-01
- Model allegedly pulled from Hugging Face after breaking benchmarks and researcher "grave concern" — djcows · 2026-10-01
- Agentic SketchUp plus generative render: talk to it and it models what you ask — snikolov · 2026-10-01