77 karma · joined April 27, 2018
Personally, the tailwind example is an argument against one specific use case: LLM-assisted/driven coding, which I also believe is the best shot of LLM being actually productive in a non-academic setting.
If I have a super-nice RL-ed (or even RLHF-ed) coding model & weights that's working for me (in whatever sense the word "working" means), and changing some function names will actually f* it up badly, then it is very not good. I hope I will never ever have to work with "programmer" that is super-reluctant to reorganize the code just to protect their pet LLM.
To them, chatgpt IS the verification.
I am not optimistic about the future. But also perhaps some amazing people will deal with the error for the rest of us, like how most people don't go and worry about floating point error, and I'm just not smart enough to see how it looks like.
Subjectively, if a student can knowledge-guide an LLM to generate response from a specific perspective, I think it is very debatable whether such usecase should be outright punished.
The part where you build a demo/prototype showcasing if an idea works, I believe it should be much easier than before as long as you postpone understanding/deciphering the implementation. In those major ecosystems there are huge collections of single-purpose libraries doing a subset of what you want to do, all you need might just be a reasonable amount of python to string them together.
However, ensuring the correctness and actually deploying them for planet-scale mandates following a reasonable subset of "best practices" which is very overwhelming, but doable. People doesn't usually build planet-scale product 10-20 years ago.