How? It's like waking up in the 90s and deciding "not to use the Internet". How would you even do it without sounding wacko? I'm not even asking "why", there are a bunch of reasons for why one may choose not to drive a car, eat beef or watch Netflix, and they may even be the right reasons, but I honestly don't understand, do you not see what's going on? The shit is out of the horse, you can't put it back, you can't pretend it's not happening; everybody understands that the horse just keeps shitting. And you're saying: "can you not make the horse shit?"...
It makes me feel very uncomfortable to say this - I don't want to appear presumptuous or patronising, but it honestly sounds almost like you have little or no clue about this.
LLMs are not about code generation. Code generation is almost like a side-effect and there are so many different things LLMs are really good at way before you get to even talk about code - gathering data by its semantic fingerprint; classification and routing where the boundary is fuzzy: is this a bug, a config error, or a support question; semantic diffing; reading stuff nobody will read: a 900-line helm file, a dependency tree, six months of changelogs, a vendor contract, etc.
If you have never dealt with cases where the input is messy and the matching criteria are impossible to write as rules, yes, of course it would feel like you don't even need LLMs. But pretty much any serious software does have plenty of those, and LLMs are incredibly good at that. Not only they are great - they are quite irreplaceable, in the sense that, yes, I won't die if I stop using the Internet, but any reason for choosing that path obviously must have some drastic rationale.
KevinMS · · focus · HN ↗
iLemming · · focus · HN ↗
KevinMS · · focus · HN ↗
iLemming · · focus · HN ↗
It makes me feel very uncomfortable to say this - I don't want to appear presumptuous or patronising, but it honestly sounds almost like you have little or no clue about this.
LLMs are not about code generation. Code generation is almost like a side-effect and there are so many different things LLMs are really good at way before you get to even talk about code - gathering data by its semantic fingerprint; classification and routing where the boundary is fuzzy: is this a bug, a config error, or a support question; semantic diffing; reading stuff nobody will read: a 900-line helm file, a dependency tree, six months of changelogs, a vendor contract, etc.
If you have never dealt with cases where the input is messy and the matching criteria are impossible to write as rules, yes, of course it would feel like you don't even need LLMs. But pretty much any serious software does have plenty of those, and LLMs are incredibly good at that. Not only they are great - they are quite irreplaceable, in the sense that, yes, I won't die if I stop using the Internet, but any reason for choosing that path obviously must have some drastic rationale.
KevinMS · · focus · HN ↗
yet that is what most people are using it for