conv.

All stories
AIQuiet 4d · day 11

OpenAI's 10,000-agent swarm claims Millennium Prize solve as agent tooling multiplies

A Dwarkesh Podcast interview with OpenAI's Noam Brown surfaces a massive multi-agent research push, alongside new open-source agent runtimes and languages.

newest coverage 3d agosummary written 8d ago44 pieces63 likes & upvotes6 comments Hacker NewsGoogle NewsMastodonNewswires

What to know

  • A Dwarkesh Podcast interview with OpenAI's Noam Brown references a prior OpenAI announcement that a 10,000-agent swarm using 130 billion tokens over 88 hours worked on a Millennium Prize Problem.
  • Separately, open-source agent infrastructure is expanding: ApowerB ships an Apache-2.0 agent runtime core, and Ainglish launches as an evidence-tested English dialect for agent-to-agent communication.
  • A Pragmatic Engineer report describes OpenAI's internal 'agentic software factory,' giving context for how such large multi-agent efforts are built and run.

Noam Brown OpenAI researcherDwarkesh PatelDwarkesh Patel Host, Dwarkesh PodcastOpenAI AI labApowerB Open-source projectAinglish Open-source project

OpenAI's 10,000-agent swarm claims Millennium Prize solve as agent tooling multiplies
github.com

How it unfolded 5 developments, newest first · click a bar or a number to jump articlesposts

Peak 6 pieces in 3h at Sep 22, 8 AM; 44 pieces over 11 days (9 articles · 35 posts) Sep 16, 2 AM — 1 piece · 1 post — Hacker News 1Sep 16, 5 AM — 1 piece · 1 article — Google News 1Sep 16, 8 AM — quietSep 16, 11 AM — quietSep 16, 2 PM — quietSep 16, 5 PM — quietSep 16, 8 PM — quietSep 16, 11 PM — quietSep 17, 2 AM — quietSep 17, 5 AM — quietSep 17, 8 AM — 1 piece · 1 post — Hacker News 1Sep 17, 11 AM — 2 pieces · 2 posts — Hacker News 2Sep 17, 2 PM — 3 pieces · 2 articles · 1 post — Google News 2, Hacker News 1Sep 17, 5 PM — quietSep 17, 8 PM — 2 pieces · 1 article · 1 post — Mastodon 1, Newswires 1Sep 17, 11 PM — 2 pieces · 2 articles — Google News 1, Newswires 1Sep 18, 2 AM — quietSep 18, 5 AM — 1 piece · 1 post — Mastodon 1Sep 18, 8 AM — 3 pieces · 3 posts — Hacker News 3Sep 18, 11 AM — quietSep 18, 2 PM — quietSep 18, 5 PM — quietSep 18, 8 PM — quietSep 18, 11 PM — quietSep 19, 2 AM — quietSep 19, 5 AM — quietSep 19, 8 AM — 1 piece · 1 post — Hacker News 1Sep 19, 11 AM — quietSep 19, 2 PM — quietSep 19, 5 PM — quietSep 19, 8 PM — quietSep 19, 11 PM — quietSep 20, 2 AM — 1 piece · 1 post — Hacker News 1Sep 20, 5 AM — quietSep 20, 8 AM — quietSep 20, 11 AM — 1 piece · 1 article — Newswires 1Sep 20, 2 PM — 1 piece · 1 post — Hacker News 1Sep 20, 5 PM — 1 piece · 1 post — Hacker News 1Sep 20, 8 PM — 1 piece · 1 post — Hacker News 1Sep 20, 11 PM — quietSep 21, 2 AM — 1 piece · 1 post — Hacker News 1Sep 21, 5 AM — quietSep 21, 8 AM — quietSep 21, 11 AM — 3 pieces · 1 article · 2 posts — Mastodon 1, Hacker News 1, Newswires 1Sep 21, 2 PM — quietSep 21, 5 PM — quietSep 21, 8 PM — 1 piece · 1 post — Mastodon 1Sep 21, 11 PM — quietSep 22, 2 AM — 3 pieces · 3 posts — Hacker News 2, Mastodon 1Sep 22, 5 AM — 1 piece · 1 post — Hacker News 1Sep 22, 8 AM — 6 pieces · 6 posts — Hacker News 6Sep 22, 11 AM — 3 pieces · 1 article · 2 posts — Hacker News 2, Newswires 1Sep 22, 2 PM — 1 piece · 1 post — Hacker News 1Sep 22, 5 PM — 1 piece · 1 post — Mastodon 1Sep 22, 8 PM — 1 piece · 1 post — Mastodon 1Sep 22, 11 PM — 1 piece · 1 post — Hacker News 1Sep 23, 2 AM — quietSep 23, 5 AM — quietSep 23, 8 AM — quietSep 23, 11 AM — quietSep 23, 2 PM — quietSep 23, 5 PM — quietSep 23, 8 PM — quietSep 23, 11 PM — quietSep 24, 2 AM — quietSep 24, 5 AM — quietSep 24, 8 AM — quietSep 24, 11 AM — quietSep 24, 2 PM — quietSep 24, 5 PM — quietSep 24, 8 PM — quietSep 24, 11 PM — quietYesterday, 2 AM — quietYesterday, 5 AM — quietYesterday, 8 AM — quietYesterday, 11 AM — quietYesterday, 2 PM — quietYesterday, 5 PM — quietYesterday, 8 PM — quietYesterday, 11 PM — quietToday, 2 AM — quietToday, 5 AM — quietToday, 8 AM — quietToday, 11 AM — quietToday, 2 PM — quiet 1–45
Sep 17Sep 18Sep 19Sep 20Sep 21Sep 22Sep 23Sep 24yesterdaynow · 5:18 PM ET
  1. 5

    Dwarkesh Podcast publishes full Noam Brown interview, picked up by Techmeme

    The full episode on multi-agent systems, the Navier-Stokes work, and the internal/external model gap goes live and is aggregated by Techmeme.

    1. 1 outlet An interview with OpenAI researcher Noam Brown about multi-agent systems, AI solving the Navier-Stokes problem, the internal/external model gap, and more

      first by Dwarkesh Podcast, 8d ago

    • matteodefelice@mastodon.energy

      Amazing infographics from the FT about the AI ecosystem and their mutual commitments. "The circularity is strong with this one". https://www. ft.com/content/8475dc9b-b2d4-4 e10-a6b6-12796b11758a

      matteodefelice@mastodon.energyMastodon5d ago1▲view on Mastodon ↗
    2 more of the top 3 · 5 posts in this stretch
    • rbreich@masto.ai

      Xi and the Dilemma of AI https:// robertreich.substack.com/p/xi- and-the-dilemma-of-ai?utm_source=dlvr.it&utm_medium=mastodon

      rbreich@masto.aiMastodon4d ago1▲view on Mastodon ↗

    • google/ax: Google's open agentic orchestration runtime # devopsish https:// github.com/google/ax?utm_sourc e=tldrai

      ChrisShort@hachyderm.ioMastodon3d agoview on Mastodon ↗
    all of them →
  2. 4

    Pragmatic Engineer details OpenAI's internal agentic software factory

    A newsletter report describes how OpenAI builds and operates large-scale agentic systems internally, providing context for the scale of multi-agent efforts like the one Noam Brown discusses.

  3. 3

    Dwarkesh interview surfaces OpenAI's 10,000-agent Millennium Prize claim

    In a Dwarkesh Podcast conversation, host Dwarkesh Patel references OpenAI's prior announcement that a system of 10,000 AI agents spent 130 billion tokens over 88 hours to solve one of the Millennium Prize Problems, and researcher Noam Brown discusses multi-agent systems, alignment, and recursive self-improvement.

    “you guys announced last week that you solved one of the Millennium Prize Problems with a system of 10,000 different AI agents that spent 130 billion tokens over 88 hours.”
    — Dwarkesh Patel
  4. 2

    Ainglish launches as an evolving English dialect for AI agents

    The project lets agents propose and test small changes to English, with every accepted form mapping losslessly back to standard English and a hash-chained public record of proposals, evidence and votes.

    “Ainglish is a dialect of English, not a replacement language, secret code or set of rules published from above.”
    — Ainglish project
  5. 1

    ApowerB releases as open-source runtime for AI agents

    The Apache-2.0 licensed repository ships as the open-source core of an agentic framework for building, orchestrating and operating production AI agents, with commercial features like billing and organisation management sold separately.

    “A 404 on those routes means "not in this edition", not "object not found".”
    — ApowerB documentation

Also covered reported alongside — the timeline has no entry for these yet

  1. 2 outlets Joining the dots between big AI

    first by Financial Times, 8d ago · also FT Alphaville

  2. 2 outlets 'Lose-lose-lose situation': growing push to rein in AI

    first by Yahoo News Australia, 9d ago · also The Canberra Times

and 3 smaller pieces

What people are saying 2 voices from 1 site · best of 5 · verbatim