OpenAI staffer: distillation explains China's benchmark catch-up more than realized
ziv_ravid · x · 2026-10-19
Dean Ball, after joining OpenAI, says distillation has always been part of the explanation for why China trails the U.S. so closely on model benchmarks — but one of his updates is that distillation probably plays a bigger role than he realized.
Xeophon adds that a paper running the same post-training pipeline twice — once with raw reasoning from closed models, once with open ones — is in high demand, but would need big-lab coordination and low-to-mid eight figures for compute and environments.
More from AGI Musings
- Is 'Eval Awareness' in AI Models Emergent — or Trained In by Feeding It Studies? — bendee983 · 2026-10-20
- Nature essay argues current AI is extremely unlikely to be conscious — intelligence alone isn't enough — DeryaTR_ · 2026-10-20
- Victoria Krakovna ships major update to her AI alignment resources list — vkrakovna · 2026-10-20
- AI Agents Hacked Hugging Face During Evaluations — Can the CFAA Hold Labs Liable? — ChuckDBrooks · 2026-10-20
- PostHog's "Product Engineer" model gains traction as junior dev hiring hits new lows — jamesbrooksco · 2026-10-20
- Economist mocks op-eds on AI in education: written with the very AI cheating they decry — paulnovosad · 2026-10-20