Noam Brown on Dwarkesh: math progress hints at what happens when AI automates research
polynoamial · x · 2026-09-18
Dwarkesh Patel released a long-form conversation with OpenAI's Noam Brown on multi-agent systems, Navier-Stokes, and what the current explosion of math progress implies once AI research itself is automated. Other key threads: how AI firms will be organized, the internal/external model capability gap, a claim that chain-of-thought is degrading, and how we'd actually know models are aligned before kicking off recursive self-improvement. Brown credits OpenAI teammates including Kevin Lu and Mike Malek for the multi-agent work. Full timestamps span multi-agent/Navier-Stokes through 'when is alignment solved'.
Related event: Noam Brown on Dwarkesh: Agent Swarms and Math Breakthroughs(3 posts)→
More from AGI Musings
- Karpathy lays out the race for a small on-device LLM "cognitive core" as JEV positions itself as a first step — Paimaamu · 2026-09-18
- Does p(doom) cover AI-assisted bioweapons, or only AI turning against us? — bosmeny · 2026-09-18
- Andrew Dai: reasoning first evolved from vision, an entire space still unexplored — AndrewDai · 2026-09-18
- Anthropic Institute, OpenAI Foundation, DeepMind Institute: do we need independent AI labs? — tobias_rees · 2026-09-18
- cspenn: So What? Defining the Agency of the Future — cspenn · 2026-09-18
- AI doomerism in retrospect: researchers now focused on surviving post-AGI, not debating if it happens — mobileraj · 2026-09-18