Hacker News Reader

Comment by daemin | Hacker News Reader

At first I thought this would be an interesting article, but as soon as they mentioned using an LLM to do the conversion I lost all interest. It's like saying "I wanted this done so I got my underling to do it, here is my story...". Like why would I bother to read it then, as it was clearly not you doing the conversion or putting any thought into it.

View on HN

The biggest problem is the author didn't even bother to verify it. I've seen big multi-model ralph wiggum or whatevers do a conversion. Run for 6 hours. Upon manual inspection to understand how it handled some tricky calculations / logic I find stubs and hard coded truthy returns. So even if you ran a smoke test suite against it -- you'd think it successful...

The biggest problem is that this is software development going forward.

Programming languages don't really matter outside craft programming.

As LLM improve, it will eventually be a matter to specify in which kind of language the specification should be generated.

The UML and RUP crowd has had their vengeance.

It doesn't have to be, we have the power to change it by just not using LLMs to generate code-like text sequences.

But if anyone is having their revenge it is the "idea guys", not the UML jockeys.

As long as upper management is measuring AI tools at work not much power is left.

The whole idea with UML jockeys is that we no longer need brick layers, that is the whole point of software factory concepts, and stuff like Rational Unified Process.

The only difference is using markdown this time around.

That does not explain why people think their project that is 99% written by LLM is worth sharing.

Post ain't about project written in LLM (and Rust one have 6 months in good ol' craft code) but about contrast between two variants.

While personally I'm skeptical about LLM usage (and vibe coding in general) I'm not going to pretend it doesn't exist just on principle. I wouldn't risk rewrite 6 months worth of code to Rails but it's a perfectly good case for LLM conversion.

The point isn't about writing it but seeing difference and making probably suboptimal choice.

I mentioned in other comment that I reviewed it extensively. It's not a big project so outside of a few spicy bits it's mostly a web app.

I'd say that saying "not putting thought into it" is unfair. I didn't push the button and YOLO.

I did research about the tradeoffs and consequences. Snippets of Rust code vs Rails are the real thing and testability of Rust app is something I spent 2 months on.

As LLM enthusiast would say: context matters ;)

you sound jealous :)

That's an assumption. How do you infer that from written text? I can not infer that.

To me it is super-strange that one uses AI to come to the conclusion that language x is better than language y. In the past people spent some time using both languages for a while, before reaching any conclusion. With AI it seems insta-gratification or insta-evaluation now. I am beginning to see why Google ruined its search engine for real humans - those who control AI rule the world now. People aren't even noticing this how dependent they are becoming on AI in general.

My experience tells me. You are disagreeing with me, but you don't give negative vibes in your comment. The other comment sounded too negative:

<< I lost all interest >> << why would I bother >> << putting any thought into it >>

It could have been worded neutrally and still convey the same points of disagreement. Interestingly, look at their other comments, they are all contentious.

I was also experimenting psychologically, to see what reactions I would get.

Reply on news.ycombinator.com