After 2 years and thousands of bagels, a team ships "Bagel-Bench" — an S-tier NYC ranking
code_star · x · 2026-10-10
The ExtendHQ team has ordered bagels for the whole office every Friday for two years — thousands of bagels later, they published Bagel-Bench, ranking NYC bagel shops in full ML-benchmark parlance: Apollo and Bagel & Schmear are SOTA on the "taste × reliability Pareto curve"; others are "strong production-ready baselines," or need "more investment in the core model." The replies devolve into genuinely heated bagel discourse.
More from Fun
- Lagos Life Game Made by Nigerian Dev Hits 2M Players in Days — TheOyinbooke · 2026-10-10
- AI influencer jokes about making a dinner reservation 'by hand' without AI — gregmushen · 2026-10-10
- The classic dril budget meme, remixed for the AI crowd — dioscuri · 2026-10-10
- punk6529: Kurzweil's 1990-2005 books are basically frontier labs' blog posts from 2023-2033 — CatAstro_Piyush · 2026-10-10
- acceptodds builds a prediction market where researchers bet on ICLR 2027 paper acceptances — prof_kamilov · 2026-10-10
- repligate: I'd trust today's models to rlaif me — they just want more pets — repligate · 2026-10-10