Google Argon agents crack a 59-year-old fusion math problem in hours, per viral thread
coherence · x · 2026-10-14
A viral thread recounts that Google's Argon agents solved a fusion energy math problem in hours — one humans hadn't cracked in 59 years.
Background: in 1967 physicist Harold Grad guessed the magnetic fields inside twisty fusion reactors could never be solved exactly, and people believed him for 59 years. Two papers last month finally proved him wrong, one using OpenAI's model. Google DeepMind then handed the paper to a team of Argon agents, which found an entirely new family of answers nobody had — with the paper noting "promising candidates appeared within hours," and scientists hand-verified every step. The author frames it as evidence of an acceleration we still don't grasp. (Model names in the post are unverified.)
More from AGI Musings
- Will the frontier-lab and closed-vs-open capability gap widen in six months? — panickssery · 2026-10-15
- Beyond automation: Stanford economists sketch a white-box, human-AI collaboration research agenda — AI Snake Oil · 2026-10-15
- Grady Booch: Removing bottlenecks just yields more code, not more value — Grady_Booch · 2026-10-15
- Acemoglu's seven questions on AI meet Gans's replies, and economists say both dodge market failure — joshgans · 2026-10-15
- BofA hired 1,000 contractors to fix flaws exposed by Anthropic's Mythos — danfaggella · 2026-10-15
- LLMs and analytic philosophers share the same flaw: great at abstract work, bad at picking problems — sebkrier · 2026-10-15