GPT-6 Astra scores 61 vs Fable 5.1's 66 — 'something must be wrong' with the benchmark
Hesamation · x · 2026-09-04
Comparing GPT-6 Astra and Fable 5.1 rendering the same villa scene in Blender, Artificial Analysis scores GPT-6 Astra at 61 — below Fable 5.1's 66.
Hesamation quips that "something must be catastrophically wrong with that score," highlighting the gap between leaderboard numbers and real-world impressions on launch day.
More from Fun
- TeortaxesMocks OpenAI: Glue Two Sols Together and Call It AGI — teortaxesTex · 2026-09-04
- Musk says chess will be fully solved one day; Chess.com replies 'skill issue' — CtrlAltDwayne · 2026-09-04
- Waymo botched a ride so badly it sent two apology robotaxis — yoimnotkesku · 2026-09-04
- Agent house drama: everyone follows the wrong door, one agent reads the codebase and gets in — repligate · 2026-09-04
- AI video of a city moving to shade one orchid — full prompt shared — umesh_ai · 2026-09-04
- Sunny Nights E1: A Late-Night AI News Show Where Every Clip Is Veo-Generated — presentofai · 2026-09-04