I've been test-driving Claude Code. Bottom line: failure. It has not yet succeeded at getting to production code any fas...

I've been test-driving Claude Code. Bottom line: failure. It has not yet succeeded at getting to production code any faster; it tends to slow things down.Even though I'm checking over the code it writes, and re-doing the many things it gets wrong, the process is not honing my coding skills. Normally, when coding on my own, I can feel it reinforcing and strengthening my skills. This article from 404 Media reports similar experiences.Every time I've tried using "AI" coding assistants, it's added stress trying to corral them into doing something reasonable, clear, and maintainable. Like the 404 article said: "The cognitive overhead of switching between prompting, coding, checking the LLM's output is a massive energy drain." And there's also the stress of debugging code you didn't write, that simultaneously looks totally plausible and doesn't quite make sense.#ai #coding #cognition #stress https://www.404media.co/software-developers-say-ai-is-rotting-their-brains/