"AMD’s AI director reports that Claude Code has become “dumber and lazier” since February, based on analysis of 6,852 sessions and 234,760 tool calls, which is the most thorough performance review any AI has received and rather more than most human employees get."
Are there any good ways to measure agent ability? Or do we just have to go by vibes?
Did you read an LLM summary of the article? It mentions all of those topics but none of them in the way you say.
The article describes a reverse-engineering effort in Bulgaria during Soviet times, but doesn't say anything anti-communist, just that direct commerce with the West was not possible at that time. It relates that effort to the author's motivation for working on an open-source circuit analyzer, and takes a pretty strong stance against uncritically adopting technology without any attempt at understanding how it works.
Are there any good ways to measure agent ability? Or do we just have to go by vibes?
The article describes a reverse-engineering effort in Bulgaria during Soviet times, but doesn't say anything anti-communist, just that direct commerce with the West was not possible at that time. It relates that effort to the author's motivation for working on an open-source circuit analyzer, and takes a pretty strong stance against uncritically adopting technology without any attempt at understanding how it works.
quote please
> generously sprinkled with anti-communism
quote please