Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Ask HN: Anyone struggling to get value out of coding LLMs?

news.ycombinator.com

Post reply on HN
Print view View original

41–50 of 293 posts

Re: Ask HN: Anyone struggling to get value out of coding LLMs?

#41

Post by Philip-J-Fry »

gyomu wrote: post #13

There are two kinds of engineers. Those who can’t stop raving about how much of a superpower LLMs are for coding, how it’s made them 100x more productive, and is unlocking things they could’ve never done before. And those who, like you, find it to be an extremely finicky process that requires extreme amount of coddling to get average results at best. The only thing I don’t understand is why people from the former gro…

I find LLMs 100x more productive for greenfield work.

If I want to create a React app with X amount of pages, some Redux stores, Auth, etc. then it can smash that out in minutes. I can say "now add X" and it'll do it. Generally with good results.

But when it comes to maintaining existing systems, or adding more complicated features, or needing to know business domain details, a LLM is usually not that great for me. They're still great as a code suggestion tool, finishing lines and functions. But as far as delivering whole features, they're pretty useless once you get past the easy stuff. And you'll spend as much time directing the LLM to do this kind of this as you would just writing it yourself.

What I tend to do is write stubbed out code in the design I like, then I'll get an LLM to just fill in the gaps.

These people who say LLMs make them 100x more productive probably are only working on greenfield stuff and haven't got to the hard bit yet.

Like everyone says, the first 90% is the easy bit. The last 10% is where you'll spend most of your time, and I don't see LLMs doing the hard bit that well currently.

Re: Ask HN: Anyone struggling to get value out of coding LLMs?

#42

Post by therealmarv »

There are two key points which are important to get most out of LLM coding assistants:

1. Use a high quality model with big context windows via API (I recommend Openrouter). E.g. Google Gemini 2.5 Pro is one of the best which keeps constant good quality (OpenAI reasoning models can be better in problem solving but it's kinda a mixed bag). Other people swear by the Claude Sonnet models.

2. Upgrade your code tools you combine with this high quality models. Google Jules and OpenAI Codex are so brand new and have a totally different aim than Cursor. Don't use them (yet). Maybe they will get good enough in future. I would focus on established tools like aider (steepest learning curve), roo code (easier) to be paired with Openrouter and if you want to have it really easy claude code (only useful with a 100-200 USD Anthropic subscription IMHO). On average you will get better results with Aider, roo, claude code than with Cursor or Windsurf.

Btw. I think Cursor and Windsurf are great as a starter because you buy a subscription with 15-20 USD and are set. It can be most likely that the more high quality tools burn more tokens and you spent more per month but you also get better quality back in return.

Last but not least and can be applied to every coding assistant: Improve your coding prompts (be more specific in regards to files or sources), do smaller and more iterations until reaching your final result.

Re: Ask HN: Anyone struggling to get value out of coding LLMs?

#43

Post by nottorp »

So you want to use a LLM on a code base. You have to feed it your code base as part of the prompt, which is limited in size.

I don't suppose there's any solution where you can somehow further train a LLM on your code base to make it become part of the neural net and not part of the prompt?

This could be useful on a large ish code base for helping with onboarding at the least.

Of course you'd have to do both the running and training locally, so there's no incentive for the LLM peddlers to offer that...

Re: Ask HN: Anyone struggling to get value out of coding LLMs?

#44

Post by namaria »

melvinroest wrote: post #37
gyomu wrote: post #13

There are two kinds of engineers. Those who can’t stop raving about how much of a superpower LLMs are for coding, how it’s made them 100x more productive, and is unlocking things they could’ve never done before. And those who, like you, find it to be an extremely finicky process that requires extreme amount of coddling to get average results at best. The only thing I don’t understand is why people from the former gro…

I'm learning math through Math Academy, ChatGPT and YouTube (mostly 3Blue1Brown and similar channels). Without that specifc combination, it would've been hell. Now it's just nice. A few years I did it with something similar to Math Academy (from the Universiteit of Amsterdam). ChatGPT wasn't intelligent enough back then so I didn't use it. It felt a lot tougher. ChatGPT answers questions that a teacher would find obs…

> So I ask ChatGPT and one of the things it said is that historically it meant "measure"

This is a great example of why using LLMs to learn is perilous. This is nonsense.

Re: Ask HN: Anyone struggling to get value out of coding LLMs?

#45

Post by prmph »

therealmarv wrote: post #42

There are two key points which are important to get most out of LLM coding assistants: 1. Use a high quality model with big context windows via API (I recommend Openrouter). E.g. Google Gemini 2.5 Pro is one of the best which keeps constant good quality (OpenAI reasoning models can be better in problem solving but it's kinda a mixed bag). Other people swear by the Claude Sonnet models. 2. Upgrade your code tools you…

It's not about the tools. The complaint here is a general one that shows up no matter what tools you are using

Re: Ask HN: Anyone struggling to get value out of coding LLMs?

#46

Post by xiphias2 »

For me anything before Codex model with webui was time waster.

Now with webui what's important is to constantly add tests around the code base, also if it gets stuck, go through the logs and understand why.

It's more of a management role of ,,unblocking'' the LLM if it gets stuck and working with it than fitting it to my previous workflow.

Re: Ask HN: Anyone struggling to get value out of coding LLMs?

#47

Post by sunrunner »

lm28469 wrote: post #18
gyomu wrote: post #13

There are two kinds of engineers. Those who can’t stop raving about how much of a superpower LLMs are for coding, how it’s made them 100x more productive, and is unlocking things they could’ve never done before. And those who, like you, find it to be an extremely finicky process that requires extreme amount of coddling to get average results at best. The only thing I don’t understand is why people from the former gro…

That's because they fall under the "engineers" category like my local postman and Michael Schumacher both fall under the "driver" category. Most people don't work on anything technically hard, most problems are business logic issues that aren't solved technically or legacy code workarounds for which you need to put 3-10 domain experts in a room for a few hours to solve.

There are a lot of terms in software development that have been co-opted from other disciplines and misrepresent a lot of development work, including 'engineering' and 'architecture'.

I think it's helpful to think of engineering as a _process_ instead of a role, and the reality is that a lot of development work doesn't necessarily rely on the strong engineering methodology (e.g. measurement, material properties, tolerances, modelling, etc.) that the people developing the software might imagine just based on the number of job adverts for 'engineers'.

This isn't a bad thing. There are hundreds or thousands of different but equally valid solutions to getting a program to do a thing, but not recognising that most code writing is somewhere between a art and engineering and is neither a purely artistic discipline but also rarely a purely engineering one is useful.

The kinds of engineering and architecture that people think of in software development only really represent common practices and shared language (e.g. design patterns, architectural patterns) and not a strong engineering practice or any kind of truth about how software actually runs.

(Unless you're writing software for launching rockets, in which case the engineering _process_ porbably should be strong).

Re: Ask HN: Anyone struggling to get value out of coding LLMs?

#48

Post by elzbardico »

melvinroest wrote: post #37
gyomu wrote: post #13

There are two kinds of engineers. Those who can’t stop raving about how much of a superpower LLMs are for coding, how it’s made them 100x more productive, and is unlocking things they could’ve never done before. And those who, like you, find it to be an extremely finicky process that requires extreme amount of coddling to get average results at best. The only thing I don’t understand is why people from the former gro…

I'm learning math through Math Academy, ChatGPT and YouTube (mostly 3Blue1Brown and similar channels). Without that specifc combination, it would've been hell. Now it's just nice. A few years I did it with something similar to Math Academy (from the Universiteit of Amsterdam). ChatGPT wasn't intelligent enough back then so I didn't use it. It felt a lot tougher. ChatGPT answers questions that a teacher would find obs…

You'd love a very old book named Mathematics for the Million. The author was a committed Marxist and an Engineer, and due to this combination, it places a strong emphasis on demystifying concepts, the historical context of mathematical ideas and its practical application.

Of course, some of it is a bit strange for modern audience, the way lots of things in calculus are taught nowadays is completely different. If you are looking for rigorous formal proofs, you won't find them on this book.

But it is a quite funny and quirk book that gives you a different perspective on math concepts and will satisfy your hunger for knowing the people and the historical context behind it.

Re: Ask HN: Anyone struggling to get value out of coding LLMs?

#50

Post by namaria »

belZaah wrote: post #25
avalys wrote: post #19

I find LLMs very useful when I need to learn something new, e.g. a new library or API, or a new language. It is much faster to just ask the LLM how to render text in OpenGL, or whatever, than read through tons of terrible documentation. Or, if I have a bunch of repetitive boilerplate stuff that I have no good templates for or cannot borrow from another part of the codebase. But when it comes to what I consider the ac…

There’s no silver bullet. Conceptualization is the hard part, exactly as you say. Mythical man-month is an important text and should be studied more.

And also the seminal papers by Gödel, Turing, and the books on cybernetics (robotics and automation in 1950s) by Ashby and Wiener.

The limits of what can be automated have been clear for almost a century now. Yet people insist in selling ways to automate your way out of social, political, economical problems. Come to think about it, people should also read Comways's paper as well.

Post reply on HN