Researcher slams OpenAI's benchmark disclosure as "sneaky and dishonest"
davidmanheim · x · 2026-11-23
Risk researcher David Manheim, in a discussion with Boaz Barak and Dan Kokotajlo, criticized how OpenAI discloses benchmark numbers: he argues OpenAI publishes a figure, waits for Anthropic to post a comparable one, then explains why the two aren't comparable — instead of stating that upfront. He called it "a sneaky and dishonest move" and said OpenAI's track record no longer earns the benefit of the doubt.
More from Companies & People
- LinkedIn cuts data export in December: contacts' emails and titles dropped — thedealdirector · 2026-11-23
- Consultant: People are lowering standards, finding AI's 'good enough' is good enough — jasonkneen · 2026-11-23
- Granola Cofounder on 'Fishing for a Problem' and the Design Behind Its Viral Onboarding — soleio · 2026-11-23
- Tobi Lütke on Jensen: Nvidia's product 'takes rocks and makes them think' — MartinGTobias · 2026-11-23
- Tesla FSD insider: end-to-end is settled, winning hinges on data diet quality — Singularitarian · 2026-11-23
- Why OpenAI is efficiency-obsessed: serving 1.2B users could bankrupt it — VraserX · 2026-11-23