Vibe coding as a term is really annoying. At what point does a project stops being considered vibe coded? If I spent a year iterating on a design and implementation using Claude code, in a domain I’m an expert in, will that still be considered vibe coded?
I'll counter you this: I don't use AI at all, but in a way I am a vibe coder, even though I'm five years of full time work into one OSS project. I have no code review. I move fast and ship bugs. I roll forward.
I see no reason to disrespect your work from what you say, but I also see no reason that AI would be much help to you after you had been learning for a year. If you are in the loop, shouldn't this be just about the moment when your growing abilities start to easily outpace the model's fixed abilities?
I feel similar but since starting to use Claude (only 4 days ago) I get it. "Vibe coding" to me is just a new term for pair programming.
vibe coding is no-look coding, it's largely being replaced by agents that do the iteration to the point no human is involved beyond initial project description like "Build me a web browser"
>At what point does a project stops being considered vibe coded?
Good question, and by the same "token" when does it start?
Maybe if there's no possible way the creator could have written it by hand, perhaps due to almost complete illiteracy to code in any language, or something like that, it would be a reference point for "pure vibe". If the project is impressive, that's still nothing to be ashamed of. Especially if people can see the source code.
All kinds of creative people I see are mostly no dummies and it might be better than nothing for them to honestly rate their own submissions somewhere on the scale from pure vibe to pure manual?
With no stigma regardless, and let the upvotes or downvotes from there give an indication of how accurate the self-assessments are. Voting directly to Show HN could even have a different "currency" [0] to help regulate the fall of Show submissions, where a single upvote could mean something like infinitely more than zero.
I'm not disappointed by a project purely vibed by somebody like a visual artist, storyteller, or business enthusiast who has never written a line of code, as long as it is astoundingly impressive, in the league of the better projects, those I would like to take a look at.
I also see real accomplished coders guide their agents to arrive at things that wouldn't be as nice if they didn't have years of advanced manual ability beforehand.
Plus I think I'm in the vast majority and have no interest in "slop", in a way that aligns with so many kinds of people who are also turned off.
But so far, the best definition we have for slop is "we know it when we see it".
Oh, well that's all I've got, so far :)
[0] slop vs non-slop which is like pass/fail, or even a numerical rating could be on the "ballot".
I view it as do you have a full mental model of the code base.
If you do then not vibe coded.
For me, I have different levels of vibes:
Some testing/prototyping bash scripts 100% vibe coded. I have never actually read the code.
Sometimes early iterations, I am familiar with general architecture, but do not know exact file contents.
Sometimes I have gone through and practically rewrote a component from scratch either because it was too convoluted, did not have the perfect abstraction I wantet/etc.
For me the third category is not vibe coded. The first 2 are tech debt in the making.