logoalt Hacker News

8-prime • today at 11:53 AM • 3 replies • view on HN

Why use C when LLMs also write stellar rust? You're not giving up any performance and you gain a modern toolchain and a safer language.

And as a next step you could have the LLM write the code such that you can use it in multiple projects. Making the token spend even more worthwile. And then maybe you can package it for other people to use as well, since the LLM wrote such stellar code that it would be idiotic for anyone to use anything else.


Replies

pjmlp • today at 12:30 PM

Neither actually, eventually they will achieve the 4 GL vision, and generate code directly.

It is already so that all Web based no code/low code tools have pivoted to LLMs, and agentic workflows.

When doing Web faced applications, it now possible to plug a bunch of SaaS and iPaaS platforms together, with zero traditional coding.

epolanski • today at 12:28 PM

It doesn't seem to me like LLMs write stellar Rust, in fact it seems to be quite questionable and full of unsafes.

At that point, just write it in C and benefit from the faster toolchain and the more comprehensive suite of tools.

➕ show 1 reply
imtringued • today at 12:23 PM

With LLMs you can let your computer write the code and you no longer have to look at the code to get a working application.

Since you no longer have to look at the code, you no longer have to review the code either.

Since you do not have to review the code anymore, you don't have to look out for memory safety or security problems.

If you don't have to look at memory safety or security problems anymore then it's obvious that there is no difference between Rust and C.

I mean, what does Rust even give you? A type checking and borrow checking compiler that tells the LLM what it did wrong? Pah, I don't have to read compiler errors anymore so obviously that means the LLM doesn't have to do that either.