LLM trained on massive human data instinctively flags content as 'too ridiculous to be real'
DanielHillSKW · reddit · 2026-09-02
A user shared a humorous observation: an LLM, having seen an absurd amount of human-generated information, instinctively classified certain content as too ridiculous to be real. The attached image shows the model's reaction to the absurd content.
More from Fun
- Made by Hand turns game crafting into a real woodworking job, one saw cut at a time — WellsLucasSanto · 2026-10-18
- LLM one-shots a Minecraft 3x3 piston door by building its own Linux test infra — Vjeux · 2026-10-18
- Yacine jokes his datacenter KPI is gigawatt-hours: get paged if not burning enough power — yacineMTB · 2026-10-18
- Siri and Google botch the same euro-dollar conversion, screenshot shows — jdluk87 · 2026-10-18
- Codex took 3 days, Opus 5.5 nailed it in an afternoon — banned user's video protest — dotey · 2026-10-18
- Peter Diamandis-backed AI documentary 'The AI Doc' hits theaters nationwide — irinarish · 2026-10-18