'Altman Wept, for There Were No More Evals to Conquer': Benchmark Saturation Joke Circles AI Twitter
jachiam0 · x · 2026-09-04
OpenAI researcher jachiam0 revived a classic AI Twitter joke riffing on Alexander the Great: "Altman wept, for there were no more evals to conquer" — a jab at benchmark saturation following the GPT-6 launch. He notes he independently rediscovered it, crediting @theojaffee and others for beating him to it by years.
Related event: 'Altman Wept': AI Community Meme Goes Viral as Benchmarks Fall(2 posts)→
More from Fun
- Screenshots of Amazon Astra's Chain-of-Thought Output Circulate on Reddit — Tough_North7059 · 2026-09-04
- No Asian Models on FelonyBench: Incentives, Not Just Alignment? — georgejrjrjr · 2026-09-04
- Gemini 3.8 Flash subagent uses honorifics when messaging its main agent — cuijiaxun · 2026-09-04
- Ovumbra: an AI-generated browser roguelite where you play a shadow creature — TAbrodi · 2026-09-04
- No Astra preview access? You're not in the top 50,000 most important people in AI — robleclerc · 2026-09-04
- GPT-6 Astra now livestreams Pokémon FireRed on Twitch — Hemingbird · 2026-09-04