LLMs are not going to allow you to do advanced programming if you couldn't already do it by hand. The thing about LLMs is, they are a force multiplier, imperfectly, but I guess they are getting there. The overall vision (unless trivial), architecture, functionality-gaps-filling, revisions, etc. of an advanced project is not going to come from an LLM.
I personally don't think we are ever going to get to that point where I can give a simple propnmt and have an LLM generate a complex app ready to run. Think about what that would require:
1. The LLM would have to read my mind and extrapolate all the minute decisions I would make to implement the app based on the prompt.
2. Assuming the LLM can get past (1), it would have to basically be AGI to be able to implement pretty much whatever I can dream up.
3. If 2 & 3 above is somehow achieved, it would be economically very valuable, and you can bet that functionality is not going to be casually enabled in LLMs, for just anyone to use.
>LLMs are not going to allow you to do advanced programming if you couldn't already do it by hand.
This is simply false. I have empirically already seen many people get LLMs to write code they couldn't do by hand in any reasonable time frame. ("Let me take a 3 month course on Python and then get back to writing this pandas script by hand" is not a reasonable time frame when your competition can pump it out for you in 30 seconds.)
Some of these people go on to learn how to code themselves, but others are perfectly happy with treating the results they're currently getting as a black box. LLMs are a force multiplier on general productivity moreso than they are on programming ability specifically.
>[I]t would be economically very valuable, and you can bet that functionality is not going to be casually enabled in LLMs
But nothing else in the world works like this. An economically very valuable thing is, almost by definition, something with a lot of demand for it. That's like saying Excel would be so valuable if only it could do math, so of course Microsoft isn't going to put a calculator into its spreadsheet program.
I personally don't think we are ever going to get to that point where I can give a simple propnmt and have an LLM generate a complex app ready to run. Think about what that would require:
1. The LLM would have to read my mind and extrapolate all the minute decisions I would make to implement the app based on the prompt.
2. Assuming the LLM can get past (1), it would have to basically be AGI to be able to implement pretty much whatever I can dream up.
3. If 2 & 3 above is somehow achieved, it would be economically very valuable, and you can bet that functionality is not going to be casually enabled in LLMs, for just anyone to use.