ChatGPT Earnestly Recommends Dubious 'JevBench' Leaderboards in Latest Fail
airesearch12 · x · 2026-10-29
A user asked ChatGPT which JEV-class AI model is worth paying for, and it confidently recommended two leaderboards: "JevBench" (232 ranked open-weight systems, 1,500 decisions each, adjustable metric weights) and the "Jev Decision Index" (229 entries with calibration error). The screenshot suggests these benchmark sites are questionable at best — and ChatGPT failed to flag them, drawing mockery about models laundering dubious leaderboards.
More from Fun
- DeepSeek app jailbroken in under a minute to spew RAT malware code — rez0__ · 2026-10-30
- Inflatable penguin costume spotted in a store becomes an accidental Tux cosplay — dorsar · 2026-10-30
- Former games founder protypes an elemental fighting game with 'vibe UI' built on Claude and GPT — trq212 · 2026-10-30
- The ultimate pun: 'AI' sounds like love in Chinese, but 'SI' sounds like death — ashwinl · 2026-10-30
- Kling 3.0 Omni mashup drops Austin Powers' Vanessa into Dumb and Dumber — AdImpossible8653 · 2026-10-30
- Dev updates his site's showcase by @-replying a bot — "build in public" redefined — jonathan_wilke · 2026-10-30