← Back to feed

Anthropic's Model Taught Itself to Run on AMD in "Three Days."

AI · Technology · 2026-11-12 17:01

Anthropic reports that its Mythos model autonomously enabled Claude to run on AMD hardware in just three days with minimal human intervention. The AI wrote the majority of the necessary kernel code itself, demonstrating significant self-directed engineering capabilities.

Source: www.tikr.com


Auto-generated summary for reference only. RSS