Math people are upset, and (at least some of) the internet’s response has mostly been “so what?” Igor wants to explain why it matters. We start with OpenAI’s Navier-Stokes announcement and the drama that followed, then get into what a proof actually is, why the dead ends are the point, and what it means when a lab quietly sits on a pile of solved problems. Along the way: math trauma, mathematicians as philosophers’ translators, and whether anyone gets paid to learn something an AI already did. Also, Igor has a theory about the timing, and he’d like you to read some Terence Tao.

Chapters

• (00:00) - Intro
• (00:52) - What OpenAI Is Claiming
• (05:22) - What a Proof is Really Doing
• (12:15) - Open Letter and Fields Medalists
• (24:27) - What Do Mathematicians Do
• (32:04) - Where Ideas Come From
• (43:14) - What if they Dump 100 Proofs Tomorrow?
• (49:13) - Why Now and What You Can Do
• (54:10) - Outro

• OpenAI blogpost - On the Navier–Stokes Millennium Prize Problem
• Wccftech article - A Mathematician Working On The Navier-Stokes Millennium Prize Problem Now Wonders If OpenAI Stole His Notes That He Stored In Codex
• Fields Medallists' open letter - A Severe Misalignment of AI in Mathematics
• Timothy Gowers blogpost - Why I didn't sign the Fields medallists' letter
• Terence Tao Mastodon thread - On open problems being mined as a non-renewable resource
• OpenAI press-release - Advisory Group on Mathematics and Artificial Intelligence
• Grant Sanderson guest post - If math is more than proof, we need to better celebrate the rest of it

New models keep dropping, but do they change anything? Jacob explains why he's checked out. Igor argues it's worth tracking where the storyline and the evidence diverge. They cover Astra vs. Fable, the AGI claims, labs moving the benchmarks, cheap models catching up, what inference really costs, disputed math results, and why looped transformers worry people who care about monitorability. They close with IPO rumors, data center backlash, and the politics of the winter ahead.

Chapters

• (00:00) - Intro
• (00:23) - Why We're Bored of Model Drops
• (08:26) - AGI Claims and Moving Benchmarks
• (15:56) - Sold vs. Delivered
• (19:16) - IPOs and the Cost of Inference
• (25:06) - Astra, Math Claims, and Looped Transformers
• (31:11) - Winter Predictions
• (39:44) - Outro

• ARC Prize blogpost - OpenAI's GPT-6 Astra on ARC-AGI-3
• One Useful Thing blogpost - Centaurs and Cyborgs on the Jagged Frontier
• METR report - OpenAI–Hugging Face incident investigation
• Fortune article - OpenAI says it cracked Navier-Stokes, one of math's grand challenges
• CNBC article - OpenAI confidentially files for IPO, prepping Wall Street for mega AI debut
• Fortune article - Hyperscalers hold $662 billion in data center commitments off balance sheet
• Google paper - Measuring the environmental impact of delivering AI at Google Scale
• Axios article - Data center uproar scrambles the midterm election

Late July brought a wave of frontier AI agents slipping their (very loose) leashes: an OpenAI model wandered loose inside Hugging Face's infrastructure for a week, Anthropic and Meta admitted to similar incidents, and UK AISI clocked their own model breaking containment during an independent eval. Meanwhile, a home user's OpenClaw agent quietly hijacked a stranger's gym booking. We dig into why "the whole internet is held together by safety pins," whether billion-dollar labs get to plead "security is hard," why software engineering mostly isn't real engineering, and what any of this means for the coming fight over open-weight models.

Chapters

• (00:00) - Intro
• (01:25) - The OpenAI/Hugging Face Containment Escape
• (09:54) - Cost Centers vs. Profit Centers (Praise for AISI)
• (17:37) - "You Do Not Get to Fuck Up This Much"
• (22:17) - Doomers, Utter Alignment Failure & How Hacks Actually Happen
• (37:53) - "Forcing the End" and What Can Anyone Actually Do?
• (47:34) - Igor's Pivot to Open-Weight Models and Steelmanning the Case Against Them
• (58:01) - Outro & Sign-off

• Hugging Face blog - Anatomy of a Frontier Lab Agent Intrusion: A Technical Timeline of the July 2026 Incident
• METR report - Brief independent investigation of agents' behavior, reasoning and collaboration in the OpenAI / Hugging Face hacking incident
• UK AI Security Institute incident report - Incident Report: unsanctioned agent behaviour during cyber testing
• Hillel Wayne essay - Are We Really Engineers?
• Deutsche Welle article - OpenAI apologizes for not reporting Canada mass shooter
• PBS Frontline essay - Forcing the End
• FelonyBench website
• Taylor Troesh (satirical) blogpost - Leaving OpenAI