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

181–190 of 293 posts

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

#181

Post by ryandrake »

michaelrpeskin wrote: post #127
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…

A little snarky but: In my experience, the folks who are 100x more productive are multiplying 100 times a small number. I've found great success with LLMs in the research phase of coding. Last week I needed to write some domain-specific linear algebra and because of some other restrictions, I couldn't just pull in LAPACK. So I had to hand code the work (yes, I know you shouldn't hand code this kind of stuff, but it w…

To steal the joke from that vintage coffee tin: LLMs help you "do stupid things faster." If all you want to do is dump terrible code into the world as fast as possible, then you've got a great tool there. If you're looking to slowly and carefully build a quality product on a solid foundation, or create a programming work of art, LLMs are at best a marginally better autocomplete.

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

#182

Post by Brystephor »

michaelrpeskin wrote: post #127
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…

A little snarky but: In my experience, the folks who are 100x more productive are multiplying 100 times a small number. I've found great success with LLMs in the research phase of coding. Last week I needed to write some domain-specific linear algebra and because of some other restrictions, I couldn't just pull in LAPACK. So I had to hand code the work (yes, I know you shouldn't hand code this kind of stuff, but it w…

> I've found great success with LLMs in the research phase of coding.

This is what I've found it most helpful for. Typically I want an example specific to my scenario and use an LLM to generate the scenario that I ask questions about. It helps me go from understanding a process at a high level, to learning more about what components are involved at a lower level which let's me then go do more research on those components elsewhere.

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

#183

Post by layer8 »

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…

> why is the letter m used when talked about an angle?

This isn’t the best example, because a simple Google search gives you immediate results for that as well, and arguably more nuanced ones. You’ll learn, for example, that m is also the standard letter for slope, which may or may not be related.

Also, if a teacher finds that question “obscure nonsense”, I’d argue that they are a bad teacher.

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

#184

Post by Vaslo »

For the mediocre coder like myself, it’s a game changer. Instead of finicking around with reading a weird dictionary schema in Python, I simply give a sample and boom I have code. Not to mention it shows me new ways of doing things that I was only familiar with and now does it by example, and then I have a better understand of using it. I’m talking about things like decorators and itertools.

I think it’s great for coders of all levels, but jr programmers will get lost once the llm inevitably hallucinates and the expert will get gains, but not like those who are in the middle.

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

#185

Post by abdullin »

I heard an interesting story from an architect at a large software consultancy. They are using AI in their teams to manage legacy codebases in multiple languages.

TLDR; it works for a codebase of 1M LoC. AI writes code a lot faster, completing tasks in days instead of sprints. Tasks can be parallelized. People code less, but they need to think more often.

(1) Maintain clear and structured architecture documentation (README, DDD context/module descriptions files, AGENTS-MD).

(2) Create detailed implementation plans first - explicitly mapping dependencies, tests, and potential challenges.

(3) Treat the implementation plan as a single source of truth until execution finishes. Review it manually and with LLM-assistance to detect logical inconsistencies. Plan is easier to change, than a scattered diff.

(4) In complex cases - instruct AI agents about relevant documents and contexts before starting tasks.

(5) Approve implementation plans before allowing AI to write code

(6) Results are better if code agent can launch automated full-stack tests and review their outputs in the process.

The same works for me in smaller projects. Less ceremony is needed there.

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

#186

Post by kragen »

lordkrandel wrote: post #119
kragen wrote: post #76

Earlier quoted context omitted.

I think you probably need a multimodal LLM to align a table with a monospace font. Blind human programmers using screenreaders will also have difficulty with that particular task. It doesn't mean they're bad at programming; it just means they can't see.

Human blind programmers may write programs to align their own tables like requested, in autonomy. Instead of me seeing them doing wrong, asking for a fix, and them doing it wrong again in an endless cycle. Also, when I tell them the solution and ask to add a feature - bam, the table is made wrong again.

True.

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

#188

Post by SpaceNugget »

codr7 wrote: post #162
l33tman wrote: post #146

Earlier quoted context omitted.

I can see why you would chime in to say that in your experience you don't get any value out of it, but to chime in to say that the millions of people who do are "inexperienced" is pretty offensive. In the hands of skilled developers these tools are a complete gamechanger.

Why? What does it add that you couldn't have written yourself faster if you're so skilled?

> if you're so skilled?

I think this is needlessly snarky and also presupposes something that wasn't said. No one said it can write something that the developer couldn't write (faster) themselves. Tab complete and refactoring tools in your IDE/editor don't do anything you can't write on your own but it's hard to argue that they don't increase productivity.

I have only used cline for about a week, but honestly I find it useful in a (imo badly organized) codebase at work as an auto-grepper. Just asking it "Where does the check for X take place" where there's tons of inheritance and auto-constructor magic in a codebase I rarely touch, it does a pretty good job of showing me the flow of logic.

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

#189

Post by sanderjd »

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…

Nah, there's a third kind, which is probably the quieter majority. These are people who find these to be useful but imperfect tools, whose expectations remain realistic, and for whom the continuing progress these tools are making is pleasant and not exactly surprising but also not taken as a given.

I'm finding that these tools are supporting an ever expanding portion of my work, while remaining frustrating frequently.

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

#190

Post by zrobotics »

ChrisClark wrote: post #172
igor47 wrote: post #149

Earlier quoted context omitted.

> For example, using an LLM to help you write a Dockerfile when you write Dockerfiles once a project and don't have a dedicated expert like a Deployment Engineer in your company is fantastic. This made me think of gell-mann amnesia: https://en.m.wikipedia.org/wiki/Gell-Mann_amnesia_effect Basically, when you understand the problem domain, you know the llm is generating crap that, as the OP says, you're unwilling to p…

As long as it's a simple enough problem and it works, I'm happy. The other day I needed to do something in powershell, which I probably use twice a year. But an LLM was just able to spit out what I needed for me. I find it's the best for that but once things get complicated, you're just arguing with the LLM, when you should have done it yourself.

I'm pretty sure I've commented this before, but it's just too ridiculous not to share again. About a year ago I needed to convert ~25k png images to webp. I happened to have them on a Windows machine, and I don't know powershell at all. I figured this would be a good thing for AI to so I asked the free version of chatGPT to write me a script to do the conversion. I was expecting a one-liner script, since I know this would be a simple bash script and there are command line tools available for this conversion on windows.

I must say, the solution was very creative. It involved using powershell to script autohotkey to open photoshop and use AHK to automate exporting each image as a webp, closing photoshop after exporting each image. I don't have a photoshop liscence, and I don't know why powershell would be needed to script another scripting tool. I also would suspect that photoshop or another Adobe tool probably has a bulk converter. But I do need to give points for originality, that is exactly the type of harebrained solution I would sarcastically suggest if I was wanting to mildly troll a coworker.

Post reply on HN