All editions

1. November 10, 2026 · 13 min read

   Edition 067 — Anthropic takes the top of the index, Meta sells a frontier-class model for 14 cents a task, and a thousand agents finish Fermat's proof in 30 hours

   Claude Fable 5.2 takes first place on the Artificial Analysis index at 65 points, Meta's Muse Ember sells frontier-class routing at about 14 cents a task, and a thousand parallel agents complete the first machine-verified proof of Fermat's Last Theorem — alongside voice cloning in a browser tab, four robots, a video agent that writes its own 3D scene, and the chip and market shocks underneath.

   • artificial-analysis
   • audio models
   • edge-ai
   • AI business
   • AI Research
   • Large Language Models
   • Humanoid Robots

2. November 6, 2026 · 12 min read

   Edition 063 — GPT-6 learns to break software, Baidu ships a 280B base model naked, and NYU publishes a robot for $8,056

   OpenAI's GPT-6 Cyber finds 81 new Firefox flaws and 948 working Linux exploits, Baidu releases a 280B-parameter base model with no instruction tuning, and Google cuts video generation to $0.07 a second. Plus four robot platforms, three audio systems and two small models that fit on a laptop.

   • Open Weights
   • Robotics
   • Video Generation
   • Frontier Models
   • Audio
   • Benchmarks
   • Small Models
   • AI business

Subscribe for each new edition →