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

I've been using Claude Code for a couple of days

twitter.com

Post reply on HN
Print view View original

361–370 of 507 posts

Re: I've been using Claude Code for a couple of days

#361

Post by theshrike79 »

InvertedRhodium wrote: post #281
develoopest wrote: post #4

I must be the dumbest "prompt engineer" ever, each time I ask an AI to fix or even worse, create something from scratch it rarely returns the right answer and when asked for modification it will struggle even more. All the incredible performance and success stories always come from these Twitter posts, I do find value in asking simple but tedious task like a small refactor or generate commands, but this "AI takes the…

My workflow for that kind of thing goes something like this (I use Sonnet 3.7 Thinking in Cursor): 1. 1st prompt is me describing what I want to build, what I know I want and any requirements or restrictions I'm aware of. Based on these requirements, ask a series of questions to produce a complete specification document. 2. Workshop the specification back and forward until I feel it's complete enough. 3. Ask the agen…

I need to steal the specification idea.

Cursor w/ Claude has a habit of running away on tangents instead of solving just the one problem, then I need to reject its changes and even roll back to a previous version.

With a proper specification as guideline it might stay on track a bit better.

Re: I've been using Claude Code for a couple of days

#362

Post by theshrike79 »

deergomoo wrote: post #246
matt_heimer wrote: post #94

Earlier quoted context omitted.

LLM are replacing Google for me when coding. When I want to get something implemented, let's say make a REST request in Java using a specific client library, I previously used Google to find example of using that library. Google has gotten worse (or the internet has more garbage) so finding code an example is more difficult than it used to be. Now I ask an LLM for an example. Sometimes I have to ask for a refinement…

I used the Kagi free trial when I was doing Advent of Code in a somewhat unfamiliar language (Swift) last year, as well as ChatGPT occasionally. The LLM was obviously much faster and the information was much higher density, but it had quite literally about a 20% rate of just making up APIs from my limited experiment. But I was very impressed with Kagi’s results and ended up signing up, now using it as my primary sear…

I did the Kagi trial in the fall of 2023 and tried to hobble along with the cheapest tier.

Then I got hooked by having a search engine that actually finds the stuff I need and I've been a subscriber for bit over a year now.

Wouldn't go back to Google lightly.

Re: I've been using Claude Code for a couple of days

#363

Post by cloudbonsai »

baq wrote: post #348
radlad wrote: post #303

Earlier quoted context omitted.

They aren't necessarily saying it can replace you. They're saying that even though it's expensive, it's cheaper than your time (which can be better spent on other tasks, as you point out.)

The first half is correct, but the conclusions shouldn’t be ‘we’re replicating our software engineers with Claude today’, they’re ‘our experienced engineers just 10x their productivity, we’ll never need to hire an intern’

I'm pondering where this "AI-automated programming" trend is heading.

For example: thirty years ago, FX trading was executed by a bunch of human traders. Then, computers arrived on the scene, which made all of them practically obsolete. Nowadays FX trading is executed by a collection of automated algorithms, being monitored by few quants.

My question is: is the software development in 2025 basically like what the foreign exchange was in the 2000s?

Re: I've been using Claude Code for a couple of days

#364

Post by becquerel »

999900000999 wrote: post #227
MyOutfitIsVague wrote: post #100

Earlier quoted context omitted.

Depends a lot on what context you feed it. If you have decent internal documentation and can feed it the right files, it tends to do quite well for me, often needing corrections or style fixes. To use it most effectively, you have to know the vast majority of the code that you want it to write ahead of time. It saves time typing, looking up methods, wrangling APIs appropriately, and sometimes it surprises me by worki…

I think the next step is to get it to run some automated testing against the code it produces and then fix issues accordingly. If I was a better programmer I'd be working on that solution right now.

Aider has built-in functionality for this. You can pass it a command (dotnet test or whatever) and it will autorun it after AI edits. If tests fail it can paste the output into the context for you.

Re: I've been using Claude Code for a couple of days

#365

Post by gloosx »

develoopest wrote: post #4

I must be the dumbest "prompt engineer" ever, each time I ask an AI to fix or even worse, create something from scratch it rarely returns the right answer and when asked for modification it will struggle even more. All the incredible performance and success stories always come from these Twitter posts, I do find value in asking simple but tedious task like a small refactor or generate commands, but this "AI takes the…

From the creators of static open-source marketing benchmarks: twitter PR posts.

Re: I've been using Claude Code for a couple of days

#366

Post by bufferoverflow »

CGamesPlay wrote: post #255

I decided to check this out after seeing the discussion here. I had previously misunderstood that it required a Claude.ai plan, but it actually just uses your API keys. I did a comparison between Claude Code and Aider (my normal go-to): I asked it to do clone a minor feature in my existing app with some minor modifications (specifically, a new global keyboard shortcut in a Swift app). Claude Code spent about 60 secon…

Now, the billion dollar question is - how long would it take you to code that diff?

Re: I've been using Claude Code for a couple of days

#367

Post by bufferoverflow »

babyent wrote: post #186
raylad wrote: post #115

I tried it on a small Django app and was not impressed in the end. It looks like it’s doing a lot, and at first I was very impressed, but after a while I realized that when it ran into a problem it kept on trying nonworking strategies even though it had tried them before and I had added to claude.md instructions to keep track of strategies and not reuse failing ones. It was able to make a little progress, but not get…

Now take your $12 and multiply it by 100k people or more trying it. Even if you won’t use it again, that’s booked revenue for the next fundraise!

That's revenue, not profit. GPUs aren't cheap.

Re: I've been using Claude Code for a couple of days

#368

Post by exikyut »

macNchz wrote: post #103
mpalmer wrote: post #84

Earlier quoted context omitted.

Reality television plus cooking show, exactly.

Cooking shows are a perfect analogy for this stuff. For some reason I never connected the highly-edited-mass-appeal "watch someone do skilled work" videos on YouTube with Food Network style content until just now, but you're right they're totally scratching the same basic itch. They make people feel like they're learning something just by watching, while there is really no substitute for actually just doing the thing…

And the sad/crazy thing is that AI assisted coding/refactoring does seem to stray substantially over the line of observation vs participation to the point that the experience is largely defined by "learning by watching"

:/

Prompt engineering (to the extent that it results in the difference between applied LLM success vs failure) requires the human to truly grok what they're getting help on - the current top comment notes that the model applied more or less exactly the same diff they would have done, based on deep understanding of the exact codebase being modified - and also have some level of intuition about how LLMs work.

Someone just treating the model and context like an obscure TV remote and pressing all the buttons to see what they do, will get a result, and it might seem interesting, but will it be ontologically correct? Good question!

Re: I've been using Claude Code for a couple of days

#369

Post by cheema33 »

smallerfish wrote: post #45
develoopest wrote: post #4

I must be the dumbest "prompt engineer" ever, each time I ask an AI to fix or even worse, create something from scratch it rarely returns the right answer and when asked for modification it will struggle even more. All the incredible performance and success stories always come from these Twitter posts, I do find value in asking simple but tedious task like a small refactor or generate commands, but this "AI takes the…

I've done code interviews with hundreds of candidates recently. The difference between those who are using LLMs effectively and those who are not is stark. I honestly think engineers who think like OP are going to get left behind. Take a weekend to work on getting your head around this by building a personal project (or learning a new language). A few things to note: a) Use the "Projects" feature in Claude web. The c…

> The difference between those who are using LLMs effectively and those who are not is stark.

Same here. Most candidates I interviewed said they did not use AI for development work. And it showed. These guys were not well informed on modern tooling and frameworks. Many of them seemed stuck in/comfortable with their old way of doing things and resistant to learning anything new.

I even hired a couple of them, thinking that they could probably pick up these skills. That did not happen. I learned my lesson.

Re: I've been using Claude Code for a couple of days

#370

Post by rhubarbtree »

develoopest wrote: post #4

I must be the dumbest "prompt engineer" ever, each time I ask an AI to fix or even worse, create something from scratch it rarely returns the right answer and when asked for modification it will struggle even more. All the incredible performance and success stories always come from these Twitter posts, I do find value in asking simple but tedious task like a small refactor or generate commands, but this "AI takes the…

I spent a few hours trying cursor. I was impressed at first, I liked the feel of it and I tried to vibe code, whatever that means.

I tried to get it to build a very simple version of an app I’ve been working on. But the basics didn’t work, and as I got it to fix some functionality other stuff broke. It repeatedly nuked the entire web app, then rolled back again and again. It tried quick and dirty solutions that would lead to dead ends in just a few more features. No sense of elegance or foundational abstractions.

The code it produced was actually OK, and I could have fixed the bugs given enough time, but overall the results were far inferior to every programmer I’ve ever worked with.

On the design side, the app was ugly as hell and I couldn’t get it to fix that at all.

Autocomplete on a local level seems far more useful.

Post reply on HN