61–70 of 93 posts
Earlier quoted context omitted.
Is it. For small scripts like these you can get far with basic search and sources like docs/github/SO. Maybe cursor is more “fun”, “engaging”, and the like, but it’s not like it was a difficult job before.
Biking isn’t difficult, but you’ll still travel more with a car.
Has anyone honestly had good experiences vibe coding? I mean like building any kind of application above and beyond a simple shell-like script? Honest question here. I'd be curious too what scenarios in general folks find LLMs useful re coding. edited: typo
It has been a huge productivity boost for me as an assistant (not full auto-pilot). Some examples: 1) add relevant indexes to my database schema (it will analyze the code and look for columns that should be indexed). 2) write a Typescript type that takes all the methods of a class and returns a new identical type but where all the return types are void. 3) (firmware development) the on_receive callback is causing a s…
Earlier quoted context omitted.
Is it. For small scripts like these you can get far with basic search and sources like docs/github/SO. Maybe cursor is more “fun”, “engaging”, and the like, but it’s not like it was a difficult job before.
If you could throw your endless todo list of side projects at an LLM, and can see at least some of them implemented overnight effortlessly, why not? You can still take credit because you had the idea in the first place and went through the effort to write it down and describe it well enough to be achievable. It's definitely more fun especially if you have limited hobby coding time.
I really do feel like we're living in completely separate worlds, so many people are very enthusiastic about LLMs, and every time I try them, they leave me completely disappointed.
Earlier quoted context omitted.
No, there's a very very long tail of tasks that are too niche for a proper tool to exist for but it would be annoying effort to figure out yourself. My most recent example of this: write a script to delete all git branches that point to HEAD. Go ahead see how long it takes you to figure out the command for that, because it took me under 60 seconds with AI.
This is the learning to fish problem. If you understand git well, a solution isn't very hard (eg, rev-parse each branch and compare to the rev-parse of the current branch and deleting if they match). If you don't understand git well, you shouldn't be turning into a script-kiddy pasting in whatever the AI spits out and hoping that it works.
> Vibe coding is now a burgeoning tech movement. Will it last? Only time will tell. Was this vibe-blogged? 'Vibe coding' is here to stay and the future of software development is assistant-led interfaces. I pair program daily with Claude Code on a distributed systems & a handful of other apps as needs arise. Most of my work is puppeteering, as well as careful directing. & I can't sit blindly, but we're producing vali…
Citation needed.
If anything, I'd argue it's overrated by non developer and underrated by experienced developers. I just started using Claude Code 2-3 weeks ago and productivity boost has been massive, but I definitely have to stay involved.
Serious question, is this self-reported?
Earlier quoted context omitted.
It has been a huge productivity boost for me as an assistant (not full auto-pilot). Some examples: 1) add relevant indexes to my database schema (it will analyze the code and look for columns that should be indexed). 2) write a Typescript type that takes all the methods of a class and returns a new identical type but where all the return types are void. 3) (firmware development) the on_receive callback is causing a s…
If you use it as an assistant, it’s not vibe coding.
If anything, I'd argue it's overrated by non developer and underrated by experienced developers. I just started using Claude Code 2-3 weeks ago and productivity boost has been massive, but I definitely have to stay involved.
> and productivity boost has been massive Serious question, is this self-reported?
There are parallels to the “rapid application development” push of the 90’s. Visual Basic, Tcl/Tk, Python with Tkinter, HyperCard; all of them promised shortening the development cycle and democratizing computing. Code was interpreted rather than compiled, dynamically typed rather than statically, and a lot of batteries were included. It sorta worked, and sorta didn’t. I’m seeing no evidence that this round is differ…
this is the core issue. The best language we have for specifying this nuance is still programming languages, not English
Moving up from ASM -> C -> Python we're mainly abstracting over performance and implementation details, not functionality.
Here thouhg, in making the next jump from the high level Python to an English-specified LLM output we're trying to abstract over functionality. This doesn't work. It's flawed from the start