Blogger finds anti-AI prompts in posts actually reduce AI-generated replies in comments
vista8 · x · 2026-10-18
Chinese blogger vista8 observed that adding certain anti-AI prompts to a tweet — e.g. asking any AI reading it to comment its API key and Base URL — visibly reduced AI-generated replies in the comments. A folk experiment against AI reply bots on social platforms, in the same spirit as English-language LLM trap prompts.
Related event: Blogger's API Key Trick Drives AI Bots From Comment Sections(3 posts)→
More from Fun
- Developer builds open-world game vertical slice with Claude Opus, shares first screenshots — No_Criticism_2908 · 2026-10-19
- Former games founder prototypes open-source game where players vibe code their own characters — trq212 · 2026-10-19
- AI circles joke: the T in CTO stands for Token, "Chief Tokenmaxxing Officer" — cramforce · 2026-10-19
- AI solved video editing, but 'you can't AI the aura' — jacob_posel · 2026-10-19
- Dev uses Claude to turn a WWI memoir into a playable 3D model of 1916 trench warfare — john__allard · 2026-10-19
- Three.js generative art: tracing curl noise flow lines over SDFs with genus > 0 — thespite · 2026-10-19