←back to thread

Generative AI coding tools and agents do not work for me

(blog.miguelgrinberg.com)
399 points nomdep | 1 comments | 17 Jun 25 00:33 UTC | HN request time: 0.209s | source
Show context
andrewstuart ◴[17 Jun 25 04:32 UTC] No.44295792[source]▶
He’s saying it’s not faster because he needs to impose his human analysis on it which is slow.

That’s fine, but it’s an arbitrary constraint he chooses, and it’s wrong to say AI is not faster. It is. He just won’t let it be faster.

Some won’t like to hear this, but no-one reviews the machine code that a compiler outputs. That’s the future, like it or not.

You can’t say compilers are slow because I add on the time I take to Analyse the machine code. That’s you being slow.

replies(2): >>44295938 #>>44297026 #
1. purerandomness ◴[17 Jun 25 08:54 UTC] No.44297026[source]▶
A compiler produces the same, deterministic output, every single time.

A stochastic parrot can never be trusted, let alone one that tweaks its model every other night.

I totally get that not all code ever written needs to be correct.

Some throw-away experiments can totally be one-shot by AI, nothing wrong with that. Depending on the industry one works in, people might be on different points of the expectation spectrum for correctness, and so their experience with LLMs vary.

It's the RAD tool discussion of the 2000s, or the "No-Code" tools debate of the last decade, all over again.