Can we stop tagging every thing as vibecoding?
39 points by LAC-Tech
39 points by LAC-Tech
Any article that even slightly mentions that AI exists is tagged as vibecoding, even when they have nothing to do with it. It:s getting out of hand.
EXAMPLE 1:
https://lobste.rs/s/ly0vif/my_students
Tagged as vibecoding, article contains a single passing reference
where people are racing to create intelligent machines, but only in order to make them slaves
EXAMPLE 2 (my own article; biased)
https://lobste.rs/s/vufbvv/user_doesn_t_care_you_should
I quoted a SINGLE example of someone putting an agentic spin on an old trope:
"Users don’t care whether the code was written by AI or by hand, or which framework you used. They care that the product works."
EXAMPLE 3
https://lobste.rs/s/gefcox/jolt_clojure_interpreter_on_janet
This is a code repo. There is no AGENTS.md, or CLAUDE.md, or any mention of AI in the readme...
EXAMPLE 4:
this post, which someone will inevitably tag as "vibecoding"
For example 3, look at the commit frequency and the size of the changes, especially on June 5th. It's too much for a human to be able to write; it's at the bare minimum 5k added lines of churn and like 2k deleted lines in a single day (eyeballed off like 10 representative commits, since there are like 50 commits that day, many of them at least 200 lines). Also, as much as I agree with the sentiment of too many things being tagged as vibecoded, I don't think example 4 is helpful to the point you want to make.
The "AI" agent files were removed 3 days ago[0]. They were there when I (and many others) flagged the post.
[0] https://github.com/jolt-lang/jolt/commit/f24c9aa1fd104ec4d36d39dc2f8908e9f0bfb787
to play devil’s advocate, example 2 clearly is a retort to vibe coding and the general state of the field, or you wouldn’t have mentioned it at all. you threw out your one reference early and it colored how I interpreted the entire rest of the post. not saying that’s good or bad, just that it happened.
How is example 2 clearly a retort to vibe coding? It might be more relevant now but I have seen these phrases uttered by many different people in various forms for years. AI is mentioned in one bullet point (the last one of four) and once openAI is mentioned more as a tech company (uttered in the same sentence as AirBnB) and that's it. To then say that an article of roughly 500 words is based on AI based on those two instances is a bit of a stretch.
Are we at the point where we're tagging things 'vibecoding' because they are perceived as a retort? If someone blogs about the importance to them of meaning and connection in writing, should that be tagged 'vibecoding' because generated slopticles lack meaning or connnection? Would it matter if it was called out as an explicit response to LLM culture / practices once, but not mentioned thereafter?
I can see the point being made about criticisms of vibecoding being lumped in with the vibecoding tag, but I'm not sure it being mentioned once, offhand should qualify.
The author for example 2 clearly states that their post is not about vibecoding. If you had read the article, you wouldn't be confused.
Example 1 is a great deal affected by the industry's buy-in to AI-generated or -assisted code, which is the site's description of vibe-coding. I suspect many posts are tagged this because they are perceived to allude to the ongoing industry movements or to using these tools, and, when people are hiding the vibe-coding tag as a way to avoid such allusions, they will inevitably tag it.
To pare down what I said elsewhere- if an article feels like an allusion, are people essentially tagging as 'vibecoding' based on vibes? I'd suggest in those circumstances the better action would be to hide the article, not tag it.
Taken to admitted extremes, time-honored advice about writing understandable code could be interpreted as a polemic against vibecoding, and tagged accordingly!
Hm, no, to be clearer, I think people tag with vibe-coding when they feel that the submission in some way is dependent on/referring to vibe-coding in a way that they don't want to engage with, and subsequently they tag it because they know others do the same. I'm speculating on other people's behaviour though so I'll just let someone who does tag aggressively answer, rather than relying on what I've heard elsewhere.
Yeah, that's a tricky thing to regulate. I can understand those who just want to be clear of vibecoding, slop, genAI, etc in all forms; but the 'in some way' that you rightly identify is doing a lot of heavy lifting for the first two examples here...
I do think the first two posts would be more appropriately tagged with culture - they do briefly reference AI (and the first in particular is clearly a reaction to it to some extent), but I personally filter the vibecoding tag because I am not interested in vibecoded projects or discussion of how to code with agents - both those posts are relevant to me and I would like to see them.
I also find it a bit annoying to see "vibecoding" put on posts about LLM security findings, use of chatbots to manage github issues, an instagram chatbot exploit, rants about hating GenAI, etc. I think it would be much more useful to have a separate tags for "writing code with LLMs" ("vibecoding") and another tag for "generative AI generally" (maybe "GenAI"?). I'd just like a bit more granularity and accuracy in the tags. I'd also like a "slop" tag for AI-written blogposts, but maybe that's a bit too difficult to enforce :)
The "slop" tag is the "spam" flag :^)
Yup; literally so, as of a recent update:
Spam which should be used for content that either is designed to promote a commercial service or for content that is created without meaningful human authorship.
(emphasis mine)
I'd like, primarily, to avoid 1) code that's vibecoded (even if it doesn't immediately qualify as "spam") and 2) discussions of workflows around vibecoded or around making software with vibecoding. if an article mentions that they vibecoded a significant element of it, i usually stop reading it, because i'm not going to learn anything if my workflow is fundamentally different, even if it's human-authored enough to have some useful technical contribution. [https://xeiaso.net/blog/2026/dancing-mad-sandboxing/](This recent post) is an example of such a post, where it doesn't seem like "spam" per se but I'm just not going to get anything out of reading it.
I can't filter these posts out though because posts like 1&2 get tagged.
I think a good metric would be "discussing the technique of vibecoding", or something like that. If something is just vibecoded like #3, that sounds like spam under the new rules, and if it's like 1 or 2, ai or culture or no tag at all sounds better.
Even with Example 3's use of AI to build it, the topic itself isn't vibecoding and I wouldn't interpret the tag to mean that. We don't tag every post mentioning ripgrep as a dev tool with "rust" because rust is not the topic of discussion.
Oh, just commented (/complained) about this on one of those stories.
I cannot really imagine a good-faith rationale for tagging these stories as vibecoding. I must admit that the cynic side of me is thinking that these votes may be coming from "pro-AI" people[1] in a sort of "protest" against the existence of the vibecoding tag at all. Like, overusing the tag such that it loses all meaning and to force its removal.
It's also possible that the tag votes are coming from the "anti-AI" side. In a sort of "you shouldn't even allude to the existence of gen-AI; get marked as a vibecoder if you do so" way.
I hope that neither of those options is the case. But i cannot really think of other reasons. If either ends up being the case: can we please stop doing this crap? It doesn't help the vibe of this place at all (pun intended). I like Lobsters. I like that communication here feels human, and not a dead internet place like other forums have become. Let's not make this place a constant flamewar about AI.
So, i would really like to know the reasons for these tag votes. IDK if suggesting a tag vote requires a reason or not. Maybe it should?
[1]: I don't like at all using this divisive language, but i want to get the point across. I normally don't use this sort of "two camps" language, because i think it play into a manufactured polarization, and it's toxic. But i'm using it here precisely to call out how toxic this polarization is, and as a maybe futile attempt to try to move the needle in the opposite direction.
I've questioned this in the past, and learned that the tag was suggested because the linked post's author had previously written a different blog post about LLM use! Some people really do want to hide anything that is even slightly connected to the topic.
Another example I was surprised got tagged as vibecoding: https://lobste.rs/s/z9lnpe/spy_interpreter_compiler_for_statically
I think the re-tagging of stories as vibecoding is similar to the recent discussion on flags. One does know who tagged it and more importantly why they tagged it because no reason can be given. In addition both flags and the vibecoding tag function as a sort of downvote for some people.