Got it — same four-way benchmark format, this time with a revolver
ChatGPT (GPT-5.6 Sol), Claude (Opus 5), Gemini (3.6 Flash), and Kimi (K3). Here's the post: Four AI models, one prompt: "generate a revolver." The physics interpretations alone tell you everything about how each
Nvidia charges $30K for an H100 with nvlink.
Tenstorrent sells the same 800GBIT/S interconnect in a $2K card. He plugged 4 into pcie x1 slots and runs OPUS-5 locally for $6/MONTH buy a tenstorrent card for $2k -> plug into any $50 pcie x1 slot -> connect cards via qsfp-dd 800g
1990 Mac Classic games - except in Color.
Opus 5 from a few weeks ago.
Most people think building an AI agent takes weeks.
This Stanford session shows one being built in 20 minutes. → 02:00 — Build your first agent with Opus 5 → 22:34 — How agents can replace entire engineering workflows → 49:47 — How builders can turn agents into serious
You have reached the end of the archive
All of Claude Opus 5