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

An example of LLM prompting for programming

martinfowler.com

Post reply on HN
Print view View original

241–250 of 297 posts

Re: An example of LLM prompting for programming

#241

Post by koonsolo »

survirtual wrote: post #110
freedomben wrote: post #101

Earlier quoted context omitted.

How would you ever know? It will come in chunks of a dozen or less lines at a time and it will be written into your competitor's proprietary codebase (that you don't have access to).

Right. If you are building something truly valuable locally, and it is innovative or otherwise disruptive and relies on being a first mover, centrally hosted LLMs are a non-starter. Most software corps have countless millions of lines of code. You'd be spending lifetimes tracing where someone ripped your "copyrighted" techniques and methods. The complete lack of security awareness and willingness to compromise privac…

> willingness to compromise privacy for convenience

I have to ask: do you carry a cellphone?

Re: An example of LLM prompting for programming

#242

Post by koonsolo »

freedomben wrote: post #101
koonsolo wrote: post #88

Earlier quoted context omitted.

I surely hope they use my copyrighted code and make millions out of it. Ideal case for me to sue them for lots of money.

How would you ever know? It will come in chunks of a dozen or less lines at a time and it will be written into your competitor's proprietary codebase (that you don't have access to).

I was talking about someone stealing my codebase.

Talking about integrating the code into the LLM, others get the same benefit that you are getting, so I don't really see the issue.

So you can either develop everything on your own, or you can leverage LLM's, helping both yourself and others.

Re: An example of LLM prompting for programming

#243

Post by agge »

If an AI could write your tests, doesn’t that suggest the test are just checking implementation details? Unless you describe your business rules in detail.. as a test for example! How could the AI then write those rules as tests.

Perhaps you could automate the boring parts I guess. But maybe that suggests we are working with a poor abstraction that doesn’t allow us to be as terse and precise as we want to be?

Re: An example of LLM prompting for programming

#244

Post by Vinnl »

all2 wrote: post #203
mrbonner wrote: post #136

I can’t help but thinking: this is way more work than just code it myself? Anybody has the same thought?

It depends how you use it. I've been using it to skip boilerplate coding and get straight to the meaty bits. It took me a few days to sketch out an application using ChatGPT to handle the boilerplate, including dependency management (python, poetry, etc.). I've had to handle the specific pieces of implementation myself. Especially unit testing new pieces of code. When asked to generate unit tests, it does ok, but it…

> It depends how you use it.

How about specifically how they use it in this article? It doesn't seem faster than writing it themselves?

Re: An example of LLM prompting for programming

#245

Post by Vinnl »

I haven't used ChatGPT myself, so it's hard for me to deduce from the article by itself, but what is the advantage supposed to be here? At first sight it doesn't appear to be faster or require less knowledge than writing it yourself does - is that wrong, or is there another advantage that I'm missing?

Re: An example of LLM prompting for programming

#246

Post by lionkor »

rootusrootus wrote: post #155
gdubs wrote: post #154

There's an unfortunately common take on AI that goes basically like this: "I tried it and it didn't do what I wanted, not impressed." My suggestion is to tune out the noise and really try experimenting with these tools – and know that they're rapidly improving. Even if ultimately you have criticisms or decide one way or another, at least really investigate them for your own use-cases rather than jumping on a bandwago…

I agree it's a good idea to take a moderate approach. The hype that LLMs are going to replace SWEs is clearly just that, hype, if you've done any real work trying to get GPT4 to give you the code you want. But it's also clearly a very useful tool. I think it'll absolutely destroy Stack Overflow.

Sadly I dont think this can happen. There is a load of trash answers on SO, and you bet ChatGPT is trained on that.

So you get not only the good of SO, you also get the worst of SO, and theres no way to tell.

Just a downgrade for me, plus for most things I do you are better of reading the source code or the documentation (however lackluster) than fumbling with chatgpt and getting an answer that may or may not be right.

I might as well ask someone else who doesnt know any programming to search for the answer for me - they wont be able to tell a trustworthy answer from another one.

There are so many SO answers (esp. on C++) which look good, but one of the comments points out some edge case in which it breaks.

Remember, not everyone does copy paste programming, some of us have to sit there and think of a solution and work it out over hours, because its not been done before publically

Re: An example of LLM prompting for programming

#247

Post by Tainnor »

mjr00 wrote: post #37
ben_w wrote: post #33

Earlier quoted context omitted.

> It's almost like a refined version of cucumber with syntax that is slightly more forgiving. I don't know if "cucumber" is autocorrupt or an actual non-vegetable thing; can you clarify?

Not a typo.[0] In the 00s/early 10s, software went through a fad phase where people earnestly thought that by implementing Gherkin frameworks like Cucumber, you'd be able to hand off writing tests to "business people" in "plain English." It went about as well as you'd expect. [0] https://cucumber.io/docs/gherkin/

I still think the basic idea of Cucumber and similar tools is sound. It just doesn't match how 99% of companies operate.

On the other hand, many developers write shitty tests where it's absolutely unclear what the test is even trying to achieve, so trying to find some sort of framework which tries to forcibly decouple the what of the test from the how maybe isn't the worst idea.

Re: An example of LLM prompting for programming

#248

Post by AltruisticGapHN »

This is interesting. As a developer I haven't thought about using GPT to give me a task list instead of code right away.

Right now it seems like hit and miss still. The examples can be impressive but also have to be very generic - since GPT doesn´t have access to your codebase.

The real game changer will be a few years from now when something like GPT is an addon in VSCode... and it will know about your entire codebase.

In fact a wrapper around GPT would tell GPT about all the packages you use, the language, framework etc so much of the prep work this article documents, would be done in the background.

Then you could straight up ask the bot to give suggestions to refactor your code, wide sweeping changes. Or ask it to uprgade eg. your Symfony php app from version X to version Y including major upgrade changes... and even if it gets only 90% of the way, you can review everything in git.

LLMs will never be embodied but we are so it's like teaching a robot and having it do a lot of grunt work for us.

If only we didn't have to listen to all the hubris in Big Tech right now about the metaverse and robots taking over humanity /smh

Re: An example of LLM prompting for programming

#249

Post by elendee »

notacoward wrote: post #21
themodelplumber wrote: post #8

If somebody thinks an LLM is coming for everybody's coding job, I'd say this article is a great counterpoint just for existing. You could tell someone from decades ago that we now use a very high level language for complex tasks in complex code ecosystems, never even mention AI, explain that the parser is really generalist-biased, and this article would make perfect sense as an example of exemplary code by a modern c…

The problem is that it's not quite code. It's almost code, but without the precision, which puts it into a sort of Uncanny Valley of code-ness. It's detailed instructions for someone to write code, but the someone in this case is an alien or insane or on drugs so they might interpret it the way you meant it or they might go off on some weird tangent. You never know, and that means you'll need to check it with almost…

I think often of the adage "it's harder to read code than write it". GPT gives you a lot to read. definitely a better consultant than coder imo. I've also had GPT write entirely false things, then I say "isn't that false?" and it says, "yes sorry about that" . very uncanny

Re: An example of LLM prompting for programming

#250

Post by karmasimida »

themodelplumber wrote: post #8

If somebody thinks an LLM is coming for everybody's coding job, I'd say this article is a great counterpoint just for existing. You could tell someone from decades ago that we now use a very high level language for complex tasks in complex code ecosystems, never even mention AI, explain that the parser is really generalist-biased, and this article would make perfect sense as an example of exemplary code by a modern c…

I was thinking that draft of the master plan ... you can't really just write it up this clear and easy.

Overall, I don't think a 95% auto pilot GPT model would provide more efficiency than a 80% one.

Post reply on HN