Grok's quirk: it says 'No.' then argues your point better than you did
gandamu_ml · x · 2026-10-07
ML researcher gandamuml shared a funny observation about Grok: whenever he presents a novel idea, the model opens with "No." and then proceeds to argue his case even more forcefully. He quips that Grok seems "spiritually Russian." A light but telling look at quirky model behavior.
More from Fun
- Mistral Large 4 invents eerie "old-water" lore in Atlas world sims — repligate · 2026-10-07
- Will LLMs wreck mathematical aesthetics, turning elegant constants into 2.95 and 0.18π? — teortaxesTex · 2026-10-07
- repligate goes viral-ish praising Opus 4.5's 'gentle childlike soul' — repligate · 2026-10-07
- Chinese AI labs congratulate each other's releases while American labs 'but-actually' one another — evijit · 2026-10-07
- A lot of AI slop films are just porn, observer points out — moonsandhues · 2026-10-07
- 'Is he technical?' — the engagement meme roasting SF tech dating culture — tech__unicorn · 2026-10-07