joerter10

I tend to agree. Unless there is some kind of major breakthrough, I see a heavy asymptotic curve in generative AI code quality

Exactly. The 'LLM coding is just like higher-level languages' analogy misses this crucial distinction.