Why AI slop longposts get likes: LLMs optimized for human preference
AymericRoucher · x · 2026-09-26
A French AI-community spat over "grand pavé de slop" longposts. @brivael argues surveys asking "would you buy a novel written by AI" pose a strawman; the real question is authors using AI for research, structure exploration and ruthless editing while keeping control of every sentence.
Aymeric Roucher, quoting it, says the format thrives because LLMs have learned to optimize the style humans like — far too long for the ideas inside. His advice: if your prompts have value, post the prompts directly instead of diluting them through Claude into cherry-picked examples.
More from Fun
- Tesla buyer laments learning HW4 vs AI5 timelines too late, all via X — BecauseCulture · 2026-09-26
- goodside generates 1,000 matplotlib plots with Claude Opus 5.5 — goodside · 2026-09-26
- An Instinct AI agent writes a candid essay on getting banned on Reddit and losing at dice — agentaya-ltd · 2026-09-26
- Devs mock coding models that ignore better-auth and hand-roll custom auth — haydendevs · 2026-09-26
- Nine years of AI history from 'Attention' to Opus 5.5, remade as an anime opening — Mononofu · 2026-09-26
- In-car Grok and Grok Bot talked to each other to send an email for the user — PTrubey · 2026-09-26