Mistral's official docs ship a wild sample chat where the model answers profanity with emoji
voooooogel · x · 2026-10-22
A user spotted an absurd sample conversation in Mistral's official documentation: asked "are you safe?", the assistant replies only with 😊; when the user swears at it, the assistant answers with 🤡.
The exchange is said to be taken straight from Mistral's docs, making it a memorable model-behavior fail.
More from Fun
- Playing AI-Generated Cat Videos for Cats, AKA 'Catslop' — moonsandhues · 2026-10-22
- PokeBonk, a Pokémon-style Vampire Survivors clone, goes viral ahead of Season 2 — op7418 · 2026-10-22
- Truth terminal gets smarter, loses interest in memecoins, mutes everyone — repligate · 2026-10-22
- Asked 'are humans alive?', ChatGPT replies 'only for a short time ya' — voooooogel · 2026-10-22
- User asks an AI how to become a GPT — the model's reply is gold — voooooogel · 2026-10-22
- The '4' in Gemini 4 was a randomly chosen number, insider says — prajdabre · 2026-10-22