Back to Latest SignalOPENAI / RESEARCH

Perplexity and Intelligence

Source: OPENAI · NOV 21, 2026

How perplexity measures progress in AI models, and why it could help developers establish.

Sources

  • Read the original at OPENAI

Around the web

  • Perplexity and Intelligence Hacker News · 89 points · 42 comments

On Bluesky

  • Yapay zekanın gerçek zekasını nasıl ölçeriz? OpenAI, modellerin gelişmiş yeteneklerini değerlendirmek için "Perplexity" (Şaşkınlık Metriği) merkezli yeni bir ortak yaklaşım öneriyor. (1/3)

    @nexlyi.bsky.social Nov 22, 2026
  • Love it when the frontier #AI labs share some of their insights. Jakub Pachocki from #OpenAI posted a big post giving more insights and transparency into how far their pretraining has progressed. #ChatGPT #Codex

    ROWAN (@gymbrowan.bsky.social) Nov 22, 2026 · 1 like
  • Perplexity and Intelligence https://openai.com/index/perplexity-and-intelligence/ (https://news.ycombinator.com/item?id=50601525)

    Hacker News 50 (@betterhn50.e-work.xyz) Nov 22, 2026
  • Perplexity and Intelligence view on hacker news

    hacker news bot (@ycombinatorbot.bsky.social) Nov 22, 2026
  • Perplexity and Intelligence #HackerNews https://openai.com/index/perplexity-and-intelligence/

    HackerNewsTop5 (@hackernewstop5.bsky.social) Nov 22, 2026

Related on Frontier

  • Anthropic Investor Says IPO Sets New Benchmark for AI Nov 19, 2026
  • OpenClaw Isn Nov 19, 2026
  • AI R&D-relevant model capabilities are improving rapidly and are getting harder to measure Nov 20, 2026
  • When low word error rate is not enough: Measuring entity accuracy in ASR Nov 20, 2026
MORE LATEST SIGNALSIGNALS