Claude computes nine-loop scattering amplitude, breaking Lance Dixon's eight-loop record
kevinweil · x · 2026-09-26
Anthropic's Science Blog reports that Claude completed a nine-loop scattering amplitude calculation in planar N=4 super-Yang-Mills, beating the eight-loop record set by SLAC's Lance Dixon and collaborators.
- The challenge was issued a month earlier by physicist-turned-science-writer Matt von Hippel, who wanted to see an LLM progress on a problem he knew was computationally hard
- Scattering amplitudes are notoriously hard: each added loop improves precision but costs exponentially more computation; most work stops at two or three loops
- Former OpenAI CPO Kevin Weil, who calculated Feynman diagrams in grad school, called the result mind-boggling and expects AI to transform high-energy physics within 12 months
More from AGI Musings
- Robotics professor on NeurIPS results and the value of 'A+B' research — GuanyaShi · 2026-09-26
- The hidden labor stack inside every price, and how AI compresses it — r0ck3t23 · 2026-09-26
- Why copper and electricity prices say little about costs after AI-driven automation — r0ck3t23 · 2026-09-26
- Allie Miller: business models aren't ready for proactive AI agents — alliekmiller · 2026-09-26
- AI PhD: CS academia is dead — ICLR submissions are just swapped-in base models — ZeeshanZiaML · 2026-09-26
- antirez: reflexive AI skeptics are just another flock — antirez · 2026-09-26