I've found codex to be better here than Claude. It has stopped many times and said hey you might be wrong. Of course this changes with a larger context.

Claude is just chirping away "You're absolutely right" and making me to turn on caps lock when I talk to it and it's not even noon yet.

i find the chirpy affirmative tone of claude to be rage inducing

This. The biggest reason I went with OpenAI this month...

My "favorite" is when it makes a mistake and then tries gaslight you into thinking it was your mistake and then confidently presents another incorrect solution.

All while having the tone of an over caffeinated intern who has only ever read medium articles.

Agree, this is really bad.

It's a fundamental failing of trying to use a statistical approximation of human language to generate code.

You can't fix it.