I've been using Claude Code for a couple of days
1–10 of 507 posts
Re: I've been using Claude Code for a couple of days
#2The second paragraph is clearly sarcastic, but the rest seems genuine, so Im a bit confused.
Re: I've been using Claude Code for a couple of days
#3I've successfully used it to fix a few issues here and there, but it also manages to make some pretty stupid mistakes. A few times it even started rewriting tests in a way so that the wrong outcome would be seen as a pass.
Re: I've been using Claude Code for a couple of days
#4I 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 wheel level" does not feel real.
Re: I've been using Claude Code for a couple of days
#5Agreed it is a step above the rest, but I find that it still needs some oversight you'd hope it doesn't. Simple stuff like rewriting stuff differently which yields the same result (to the naked eye, no need to even test it), to having to call it out when it misses things (update other scripts with the same problem, reflect the changes in the README or requirements.txt) and to ask it to try harder to solve an issue in a better way.
Sometimes it takes the easy way out. If you look up your billing, you will see sometimes it uses the 3.5 instead of the 3.7 API, maybe this has something to do with it. It apologies ("you are correct", "yes that was hasty") but I'd rather have it try harder every time, not only when called out (and foot the bill that goes with that ofc).
But overall the experience is great. I hope to use it for a few months at least until something else comes along. At the current pace, I switch subscriptions/API/tool at least once a month.
Re: I've been using Claude Code for a couple of days
#6I 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 don't believe in those either, and I never see compelling YouTube videos showing that in action.
For small stuff LLMs are actually great and often a lifesaver on legacy codebases, but that's more or less where it stops.
Re: I've been using Claude Code for a couple of days
#7I've successfully used it to fix a few issues here and there, but it also manages to make some pretty stupid mistakes. A few times it even started rewriting tests in a way so that the wrong outcome would be seen as a pass.
Kinda reminds me of how when it finds issues with typescript it hacks the types rather than refine the values or business logic.
Re: I've been using Claude Code for a couple of days
#8The second paragraph is clearly sarcastic, but the rest seems genuine, so Im a bit confused.
The second paragraph isn't sarcastic. At least not w.r.t. to Claude Code. That bit there about North Korean hackers is mild sarcasm, but has no bearing on the remainder of the post.
Re: I've been using Claude Code for a couple of days
#9It’s ok.
It’s expensive and is correct for the easy 90%. It messes up the hard 10%.
I guess with it rapidly improving, I should wait 3 months. What’s frustrating is when you spend 20$ in credits on it writing non functional code.
That said, every programmer needs to at least demo these tools
Re: I've been using Claude Code for a couple of days
#10[flagged]