> I almost never write code directly (it's always Claude)
Who, then, understands the code? If the answer is "no one really", entropy will overwhelm your codebase sooner or later. Otherwise, you need to read the code, and for that the knowledge of language is still relevant.
> If the answer is "no one really", entropy will overwhelm your codebase sooner or later. Otherwise, you need to read the code
I think about this on the regular -- I know the answer is currently "you own the code, so you have to understand it", but to unlock the true productivity multiplier, in the future, the answer has to be "no one really".
I think about it using the concepts from my job (academia) -- to actually have PhD student-level intelligence means that you have to trust that it does a good enough job that you can focus on other stuff. Professors often bring the correct ideas or intuitions, but they have to trust the PhD student to write the code and/or fill in the gaps in the proofs -- they can advise them on the high-level issues during a consultation, but that's about it.
I am pretty bad at working in the current LLM workflow -- it is tough for me to focus on reading a TCS paper for review, keeping all the details and invariants in my head, but every 5-10 minutes go to my PC, completely switch contexts/projects, read the code and think about the LLM's comments, suggest the next step, and then go back to reading.
I do, with different tools and most of the time at a higher level of abstraction. The same way we understood the machine code 2 years ago, even though we didn't write it directly. Just another layer, nothing more.
I don't think you can really understand assembler without writing it and since compilers are fairly deterministic and get constant attention most of us using them haven't truly had to fight with what turned out to be a compiler bug.
Whereas LLMS regularly produce shit. One can excuse them for not understanding one's turn of phrase or whatever but it amounts to the same problem in the end - you have to understand the output language a lot better than most people ever had to understand assembler.
It feels like a total waste of time and I wonder if other feel the same.
One of the consequences of the LLM tsunami might be the freezing of research and development in programming languages.
Maybe we'll be stuck with J's and python forever...