Pular para o conteúdo

What OpenAI Actually Did to Navier-Stokes (EP 58)

24 de setembro de 2026 · 4 h 9 min

Notas do episódio

What does it mean to solve an equation that describes almost every fluid around us, from the air over a wing to the water swirling down a drain?

In Episode 58 of From First Principles, Lester Nare and Krishna Choudhary build the Navier-Stokes equations from the ground up before digging into OpenAI’s claimed breakthrough and the debate surrounding it.

Summary

  • How Newton’s laws become equations for a moving fluid
  • Velocity fields, incompressibility, pressure and the nonlinear convective term
  • Why viscosity smooths a fluid while nonlinear motion can create finer structure
  • What finite-time blowup means, and why simulation is different from proof
  • How forced and unforced equations differ, and why those assumptions matter
  • Earlier work on Euler, Boussinesq and related fluid equations
  • OpenAI’s claimed result, Lean verification and the scope of the theorem
  • The dispute over scientific credit and the human research behind AI-assisted work
  • The METR investigation of the Hugging Face incident
  • Emergence World and long-running multi-agent experiments
  • AI-assisted biological discovery, oversight and recursive self-improvement
  • Separating demonstrated capabilities from claims and future scenarios

Chapters

00:00 Can AI solve Navier-Stokes?
00:34 Episode introduction
02:20 Navier-Stokes: Mathematics Meets AI
10:20 Building the Equations of Fluid Motion
21:34 Velocity fields, divergence and incompressibility
34:33 Acceleration and the convective term
52:18 Why Fluid Motion Is Nonlinear
1:02:08 Pressure, Euler and the Missing Physics
1:14:50 How Viscosity Changes Everything
1:33:36 Solving Equations vs. Simulating Fluids
1:44:59 Can a Smooth Fluid Blow Up?
2:13:52 The Road to the Claimed Breakthrough
2:31:13 Inside the Claimed Navier-Stokes Proof
2:48:26 The Dispute Over Scientific Credit
3:07:23 From Chatbots to Agents
3:08:57 The METR report and Hugging Face incident
3:23:23 AI Risk, Oversight and the Race Ahead
3:38:51 AI Discovery Beyond Mathematics
3:52:47 Why “just turn it off” gets complicated
4:06:47 Closing thoughts and what comes next
4:08:46 Outro

Featured Research

OpenAI’s Navier-Stokes announcement
Tristan Buckmaster’s statement
METR investigation
Emergence World
AI-assisted enzyme discovery

Explore FFP
ffppod.com
ffppod.com/funding
ffppod.com/transfers
ffppod.com/America250

Watch on YouTube
youtu.be/NGfGw1tGxUY

Support the show
ffppod.com/donate

Follow
@FFPPod on X / Instagram / TikTok / Facebook

Todos os episódios de From First Principles

Configurações

Exibição

Idioma

Define quais estações você vê primeiro na página inicial.

Histórico

As rádios e episódios ouvidos recentemente e suas últimas buscas ficam só neste navegador.

Avisos

Um aviso rápido quando um podcast que você segue tem um episódio novo. Desative podcasts individuais pelo sino na página deles.

Conta

Salve favoritos – no app e no site.

Atalhos de teclado

/
Abrir busca
Espaço
Reproduzir ou pausar
← →
No player: voltar 15 s, avançar 30 s (episódios) · na fileira: próximo bloco
↑ ↓
Na fileira: alvo dentro do bloco
Home / End
Na fileira: primeiro ou último bloco
Esc
Fechar janela ou menu
?
Esta visão geral
Tocando agora

Ir para a estação

A seguir

    Adicione episódios com “A seguir” – eles tocam depois do episódio atual.