Debate: how much did 10,000-agent parallel scaling actually contribute to the Navier-Stokes proof
scaling01 · x · 2026-09-18
A technical spat over how much credit the agent swarm deserves in the AI-assisted Navier-Stokes proof. One side claims the swarm's contribution is "probably quite low (10%)"; the counterargument is that the plot in question shows serial test-time-compute scaling, not parallel—and the proof literally wouldn't exist without this exact parallel scaling to 10,000 agents.
Two distinct questions emerge: how much did parallel TTC scaling to 10,000 agents actually speed up the discovery, and what role did the swarm itself play?
Related event: Debate Erupts Over Agent Swarm's Role in Navier-Stokes Proof(12 posts)→
More from coding & agent
- Dev hits multi-credential pitfall in NewMax Provider: one key per provider doesn't scale — yangyi · 2026-11-04
- Musk cites unsustainable Grok Bot costs as Hermes offers one-command agent migration — Teknium · 2026-11-04
- Open-source agent Cindy launches: cross-device tasks, files and model sharing — yangyi · 2026-11-04
- Multi-Agent Loop Workflow Curates 1,132 Top Bilibili Videos Into a Website — huangyun_122 · 2026-11-04
- Graphic designer builds NASA-data 3D Solar System in 2 months with Claude Code, no coding — Shevei1342 · 2026-11-04
- How to Run a 5-Person Office With Claude: Assistant, Bookkeeper, PM, Analyst & Office Manager — evielync · 2026-11-04