Fast and Hard Code • Hacker News Board

Re: Fast and Hard Code

#11

Post by applfanboysbgon » Sat, Aug 22, 2026, 10:08 PM UTC

throwaway63467 wrote: post #8

I wonder how people can like LLM generated code so much and then at the same time find LLM generated prose so repulsive. Is the quality of LLM generated code higher? I would think there’s much more prose than code to train on. Or is it that we don’t care about that as we’re only looking at the effects of that code? Maybe people pick Rust as they won’t understand what the LLM writes anyway so there’s less guilt about…

> much more prose than code to train on

This is completely irrelevant. An LLM's voice is not influenced by its at-scale training data. LLM-prose is the result of human feedback training rewarding it for making every single sentence sound like a YouTube or BuzzFeed headline, which is effective against a large portion of the population who get dopamine hits from such clickbait style of writing, and completely infuriating to the portion who recognize it for what it is.

LLM code is also infuriating, but for a different reason, because it is not reinforced in exactly the same way. However, people are fine with it because they don't read it. If they actually read the dogshit code an LLM produces, and were capable of understanding it, they would be infuriated by it too.

Re: Fast and Hard Code

#12

Post by javcasas » Sat, Aug 22, 2026, 10:16 PM UTC

patrickmay wrote: post #6

sublinear wrote: post #4

I still think LLMs generate boilerplate and examples that should have been in the docs to begin with. We never really needed LLMs. We needed more active communities and maintainers to care about the "boring" parts. All the rest of the generated code is usually slop. It only gets worse the further away you get from the most popular idioms that have already been written. I don't think we're ever getting beyond that poi…

Agreed. Whenever I hear about LLMs automating "boilerplate" I ask "Why didn't you abstract that away before LLMs existed?"

I kept saying that about heavy design patterns and unergonomic OO programming languages that end up requiring code generators (you know exactly what programming languages I'm saying). I kept hearing rejection after rejection, and angry words.

People don't like abstracting, or just don't understand abstractions, often just mistaking it for indirections. The few that understand it just take the message, apply it and go away to live their lives with improved coding.

Re: Fast and Hard Code

#13

Post by xboxnolifes » Sat, Aug 22, 2026, 10:21 PM UTC

throwaway63467 wrote: post #8

I wonder how people can like LLM generated code so much and then at the same time find LLM generated prose so repulsive. Is the quality of LLM generated code higher? I would think there’s much more prose than code to train on. Or is it that we don’t care about that as we’re only looking at the effects of that code? Maybe people pick Rust as they won’t understand what the LLM writes anyway so there’s less guilt about…

Because code isn't prose. It has a much more structured, "correct" form than prose. After adhering to a high level structure of a program and applying linter rules, there's really not that much variance between code (In terms of how it reads, not correctness or performance). LLM code without guard rails can get really bad, but with some guard rails it's generally not worse than the average code I'd read before LLMs. I'm used to code being generally bad.

Books or blogs on the other hand are different. Every person writes differently, and the writing style is part of the experience. The same scene or concept can be described in an infinite number of ways. But LLM prose ends up very samey. You get the same sentence structure, the same key words, the same metaphors, the same similes, the same rough distribution of word usage. It's very average, and it's boring. And it frequently reads like a corporate blog.

Re: Fast and Hard Code

#14

Post by CoolestBeans » Sat, Aug 22, 2026, 10:31 PM UTC

So I've been working on a project in Rust because a library that's appropriate for the design has a terrific Rust API and a miserable C API. I've been loosely tracking Rust since before 1.0 and I've never felt it's a language that's easy to be productive in, and not just because of lifetimes and the borrow checker. So I've been using coding agents to get me through. The problem is I don't have a sense of what feels good or bad, so I can't direct the AI to steer in another direction. And neither does AI it turns out. So I'm still going really really slow. AI is still not a substitute for lack of experience or knowledge.

The best thing I can compare it to is Adderall (without a prescribed reason for taking it). Have you ever been reading a dense text and realized halfway through the page that you didn't comprehend anything? I get this kinda sick feeling in my stomach and then I back up to the last point I was really paying attention. Adderall gets rid of that feeling, and I can just read through. But am I actually comprehending the text or is the drug just raising the ambient dopamine levels enough that I'm not feeling the missed comprehension? Using coding agents makes me feel the same way. I'm not feeling the struggle of learning something new, but I worry I'm writing code no one will be able to fix, even the AI.

Anyways, I've defaulted to using a free chatbot "ChatGPT style". It makes a lot of mistakes and forces me to go through everything over and over.