He vibecoded an LLM-based photo rating system to check his 5/10 self-assessment
AaronBergman18 · x · 2026-10-29
After friends disputed his self-rated 5/10 attractiveness, the author vibecoded a photo-rating pipeline using LLM scores as non-final inputs, validated against real anonymous rating data. Result: mean 5.1, median 5.2, range 3.8-6.1 — friends were just being nice.
Related event: Developer builds LLM-based attractiveness rater, scores himself 5.1(2 posts)→
More from Fun
- Live recursive self-improvement: kernel optimization loop streamed in real time — heyshrutimishra · 2026-10-29
- Running Omarchy: the 'you must broadcast you use Arch' meme lives on — granawkins · 2026-10-29
- Mac Studio Ultra 512GB upgrade prices mocked: 'Dad sold compute during the Apple silicon AI frenzy' — vincent_koc · 2026-10-29
- Costume party pun: 'Your costume's too small to be a harp' — 'Are you calling me a lyre?' — dioscuri · 2026-10-29
- Keychair Lounge: an AI-generated lounge chair upholstered in keycaps — Delahuntagram · 2026-10-29
- Void dubbed the 'Vercel of Cloudflare' as Syntax podcast covers Deno's Cloudflare acquisition — cnakazawa · 2026-10-29