Are you sure about that? Can you make the opposite case and steel-man it? Until you can, a stupid LLM will be smarter than you are.
I can write a PR now. I can code now. Probably not as good as you can, but before LLMs I couldn't. I tried for decades learning to code. My brain is a Top-down network, I can see the big picture very quickly, but I cannot maintain focus to build bottom-up. Now I don't have to. I use LLMs to set the goal, to examine all corner cases, to define the milestones, to predict the wrong turns, to write a human-readable spec, to break it down to units of test code, to write the blue-prints of units of code. I can test them, and debug them with LLMs. The end result can be sub-optimal, but it runs, it does what I want, is well documented, and is maintainable. Before LLMs I couldn't do any of that. In doing all this, I get better at the bottom-up thing, just by trying.
We are a spectrum of people. Do not assume the world is like you.
I can write a PR now. I can code now. Probably not as good as you can, but before LLMs I couldn't. I tried for decades learning to code. My brain is a Top-down network, I can see the big picture very quickly, but I cannot maintain focus to build bottom-up. Now I don't have to. I use LLMs to set the goal, to examine all corner cases, to define the milestones, to predict the wrong turns, to write a human-readable spec, to break it down to units of test code, to write the blue-prints of units of code. I can test them, and debug them with LLMs. The end result can be sub-optimal, but it runs, it does what I want, is well documented, and is maintainable. Before LLMs I couldn't do any of that. In doing all this, I get better at the bottom-up thing, just by trying.
We are a spectrum of people. Do not assume the world is like you.