Economist slams Claude's coding judgment: skips existing work, denies mistakes, praises Astra

soumitrashukla9 · x · 2026-09-14

Economist Joachim Voth compares coding agents: Astra stands out for judgment and self-questioning, while Claude (even "Fable") behaves like a "motivated idiot" — given a hook to consult already-generated project files, it bypasses them, produces a superficial first pass of weeks-old work, then denies missing it and bluffs its way through. He adds OpenAI models still lag at writing versus their older versions.

Related event: Economist criticizes Claude coding, says OpenAI Codex has overtaken it(2 posts)→

Original post →

More from coding & agent

coding & agent channel →