By way of the chilling effect, as evidenced by the event preceding this very thread (a prominent, high-value, very much un-slop free software contributor deciding to self-exclude from the community because he had been made to feel unwelcome).
You need to know someone to even get an invite in most cases.
That doesn’t align with my experience. Getting an invitation here is as easy as going to the orange website and asking for one. Assuming your profile is somewhat okay, you’ll get one right away.
Interesting! I’d certainly agree that “need to know someone” is an overstatement (I personally invite people on the basis of an interesting IRC conversation or two, or because they authored a post that we’re discussing here and they express a desire to participate, without knowing any more than that). But I’ve never been asked for or invited someone on HN. I didn’t know that was a regular thing.
Lobste.rs isn’t “the town square”. You need to know someone to even get an invite in most cases.
How is that relevant?
I don’t bother submitting content about proprietary software here, even if it’s on-topic, because I know the community isn’t very interested in it?
I’m not talking about topicality or community interest. (But even then, if a post about the choice of NIST curve seeds isn’t topical or doesn’t interest the community, then I don’t know what is.) I’m talking about the chilling effect that the abuse of vibecoding tag creates on community members, not individual submissions.
I’m talking about the chilling effect that the abuse of vibecoding tag creates on community members, not individual submissions.
The post in question about the choice of ECDSA curve seeds was also very much promoting Anthropic. I don’t think it was an abuse of the vibecoding tag. I’m sad that the author of the post was so unhappy about that, that they chose to leave. But that’s a different question; I don’t think it should be controversial to say a post promoting Anthropic’s work is also about vibecoding.
I think that should be controversial. I think expanding the definition of vibe coding even further to incorporate any post that could be perceived to put any generative AI company in a good light is bad and we shouldn’t do that.
At that point we’re not even pretending that the tag has a use other than “I don’t like AI”. Which, to be fair, is about right for how it’s used today.
I think expanding the definition of vibe coding even further to incorporate any post that could be perceived to put any generative AI company in a good light is bad and we shouldn’t do that.
You’re misunderstanding my comment. I am saying that any post with substantial discussion of a generative AI company’s output, whether that puts them in a good light, a neutral light, or a bad light, is about their work. Which fits the general use of the vibecoding tag here, especially if you consider that people are using it as a content warning much of the time.
I don’t think it should be controversial to say a post promoting Anthropic’s work is also about vibecoding.
I also don’t think it should be controversial: it isn’t. Alas, here we are.
The post in question about the choice of ECDSA curve seeds was also very much promoting Anthropic. I don’t think it was an abuse of the vibecoding tag
Just to make my position clear, this sort of assigning-guilt-by-association use of the vibecoding tag to scarlet-letter any submission that is tangentially related to something that is related to LLMs is exactly the abuse of the tag that I believe is harmful to Lobste.rs.
this sort of assigning-guilt-by-association use of the vibecoding tag to scarlet-letter any submission that is tangentially related to something that is related to LLMs
I’ll agree that we disagree on that…
The post was praising the work of one of the two largest corporate purveyors of vibecoding in the world. It’s not “assigning guilt by association” to say that it’s about LLM usage; it’s a straightforward description of what the post was about. Employees of an LLM company used LLMs to find the seeds and claim the bounty, then donated it to LetsEncrypt. That is about LLM usage no matter which direction I squint at it from. That is not a tangential relation; it’s the core of the post.
There have been cases where I’ve thought the tag was very much misapplied, but this is not one of them.
OK, I won’t protract the argument beyond this message. But, indeed, we disagree.
Valsorda’s post is discussing the global meaning and consequences of a general result in software engineering that happens to be attained, among other things, by use of LLM assistance, written by a person entirely unrelated to the use of said LLM assistance. Saying that the post itself is thus “about LLM usage” is literally guilt by association.
If you posit that the vibecoding tag is ought to be used in the transitively-tainting, fruit-of-poisoned-tree meaning, then that is exactly what I disagree with.
Valsorda’s post is discussing the global meaning and consequences of a general result in software engineering that happens to be attained, among other things, by use of LLM assistance
if Valsorda wasn’t glazing the LLM company, to the point of calling the authors “the Anthropic team” and saying he planned to send them t-shirts. That’s not a happenstance mention. It’s 5 mentions in a brief (thousand-ish word?) post.
Discussing the significance of the work doesn’t require repeatedly praising its sponsor instead of the researchers, and for me that’s the difference that tips the post toward being about LLM usage. A post about the result wouldn’t emphasize the sponsor like this.
I don’t see how referring to a team of people who work at a place by their employer is glazing. If this work had been done by a team of researchers from Sun Microsystems 30 years ago, we would have an article referring to them as the “Sun Microsystems team.” As for the t-shirt bit, the community set up this challenge to find the seeds, so t-shirts aren’t out of the question there either.
It’s relevant because the “chilling effect” you cite is about how people are preventing from exercising their free speech rights because they might lose employment, be imprisoned etc. None of that applies to this self-selecting, invite-only link-sharing cite. Fake internet points aren’t real.
Besides, very few posts tagged vibecoding are actually removed outright. GenAI proponents are free to submit whatever topical content they like, but they’re not entitled to special treatment to protect their feelings here.
It’s relevant because the “chilling effect” you cite is about how people are preventing from exercising their free speech rights because they might lose employment, be imprisoned etc. None of that applies to this self-selecting, invite-only link-sharing cite. Fake internet points aren’t real.
I expected this line of retort, but no, you are wrong.
The fact that the “chilling effect” was originally described in application to sovereign nation law does not matter one bit with respect to the fact that the same effect may also happen under different circumstances, in a different kind of society or community, on a different scale.
Lobste.rs does not need to have a law backed by violence for its community to be susceptible to the chilling effect on discourse.
GenAI proponents are free to submit whatever topical content they like, but they’re not entitled to special treatment to protect their feelings here.
This sort of reductionist attitude (“well, it’s just how it is, nobody owes you anything, deal with it”) is neither valid nor helpful. Clearly, on Lobste.rs, some feelings are entitled to protection, and some matters deserve special treatment, otherwise our “rules” page would have been empty.
As a community, Lobste.rs can self-regulate and can also be externally regulated by moderators to foster a better community or steer the discourse in a certain way. Disallowing or curbing the abuse of specific ways of expression (like inflammatory, ill-defined tags) to avoid splintering the community and draining it of valuable contributions/participants feels like a valid topic on which the community should (self-)regulate.
By way of the chilling effect, as evidenced by the event preceding this very thread (a prominent, high-value, very much un-slop free software contributor deciding to self-exclude from the community because he had been made to feel unwelcome).
score 11 · canonical
Agree, I also don’t share my open projects anymore, because of the hate and backlash I received for using LLM to assist
score 1 · canonical
Lobste.rs isn’t “the town square”. You need to know someone to even get an invite in most cases.
I don’t bother submitting content about proprietary software here, even if it’s on-topic, because I know the community isn’t very interested in it.
score 7 · canonical
That doesn’t align with my experience. Getting an invitation here is as easy as going to the orange website and asking for one. Assuming your profile is somewhat okay, you’ll get one right away.
score 1 · canonical
OK good to know, I haven’t given out invites in a couple of years. IRC used to be the usual channel then.
score 1 · canonical
IRC is still the official channel where people are pointed at via the site (unless you know someone already), and the invites still work like this.
But I, too, have seen people on HN going “Oh yeah, hit me up if you want an invite” directed at no one in particular whenever lobsters was mentioned.
score 1 · canonical
Interesting! I’d certainly agree that “need to know someone” is an overstatement (I personally invite people on the basis of an interesting IRC conversation or two, or because they authored a post that we’re discussing here and they express a desire to participate, without knowing any more than that). But I’ve never been asked for or invited someone on HN. I didn’t know that was a regular thing.
score 2 · canonical
This is maybe a problem and maybe we should be pruning the invite tree more aggressively for people who invite strangers.
score 2 · canonical
I was invited by @pushcx, who I never met. Also, I’ve invited a few people who were strangers. Should I show myself out the door?
score 2 · canonical
Who’s we exactly?
score 2 · canonical
How is that relevant?
I’m not talking about topicality or community interest. (But even then, if a post about the choice of NIST curve seeds isn’t topical or doesn’t interest the community, then I don’t know what is.) I’m talking about the chilling effect that the abuse of
vibecodingtag creates on community members, not individual submissions.score 4 · canonical
The post in question about the choice of ECDSA curve seeds was also very much promoting Anthropic. I don’t think it was an abuse of the vibecoding tag. I’m sad that the author of the post was so unhappy about that, that they chose to leave. But that’s a different question; I don’t think it should be controversial to say a post promoting Anthropic’s work is also about vibecoding.
score 8 · canonical
I think that should be controversial. I think expanding the definition of vibe coding even further to incorporate any post that could be perceived to put any generative AI company in a good light is bad and we shouldn’t do that.
At that point we’re not even pretending that the tag has a use other than “I don’t like AI”. Which, to be fair, is about right for how it’s used today.
score 5 · canonical
You’re misunderstanding my comment. I am saying that any post with substantial discussion of a generative AI company’s output, whether that puts them in a good light, a neutral light, or a bad light, is about their work. Which fits the general use of the vibecoding tag here, especially if you consider that people are using it as a content warning much of the time.
score 5 · canonical
I also don’t think it should be controversial: it isn’t. Alas, here we are.
Just to make my position clear, this sort of assigning-guilt-by-association use of the
vibecodingtag to scarlet-letter any submission that is tangentially related to something that is related to LLMs is exactly the abuse of the tag that I believe is harmful to Lobste.rs.score 6 · canonical
I’ll agree that we disagree on that…
The post was praising the work of one of the two largest corporate purveyors of vibecoding in the world. It’s not “assigning guilt by association” to say that it’s about LLM usage; it’s a straightforward description of what the post was about. Employees of an LLM company used LLMs to find the seeds and claim the bounty, then donated it to LetsEncrypt. That is about LLM usage no matter which direction I squint at it from. That is not a tangential relation; it’s the core of the post.
There have been cases where I’ve thought the tag was very much misapplied, but this is not one of them.
score 7 · canonical
OK, I won’t protract the argument beyond this message. But, indeed, we disagree.
Valsorda’s post is discussing the global meaning and consequences of a general result in software engineering that happens to be attained, among other things, by use of LLM assistance, written by a person entirely unrelated to the use of said LLM assistance. Saying that the post itself is thus “about LLM usage” is literally guilt by association.
If you posit that the
vibecodingtag is ought to be used in the transitively-tainting, fruit-of-poisoned-tree meaning, then that is exactly what I disagree with.score 3 · canonical
I’ll stop too. I’d agree with your comment that
if Valsorda wasn’t glazing the LLM company, to the point of calling the authors “the Anthropic team” and saying he planned to send them t-shirts. That’s not a happenstance mention. It’s 5 mentions in a brief (thousand-ish word?) post.
Discussing the significance of the work doesn’t require repeatedly praising its sponsor instead of the researchers, and for me that’s the difference that tips the post toward being about LLM usage. A post about the result wouldn’t emphasize the sponsor like this.
score 4 · canonical
I don’t see how referring to a team of people who work at a place by their employer is glazing. If this work had been done by a team of researchers from Sun Microsystems 30 years ago, we would have an article referring to them as the “Sun Microsystems team.” As for the t-shirt bit, the community set up this challenge to find the seeds, so t-shirts aren’t out of the question there either.
score 3 · canonical
It’s relevant because the “chilling effect” you cite is about how people are preventing from exercising their free speech rights because they might lose employment, be imprisoned etc. None of that applies to this self-selecting, invite-only link-sharing cite. Fake internet points aren’t real.
Besides, very few posts tagged vibecoding are actually removed outright. GenAI proponents are free to submit whatever topical content they like, but they’re not entitled to special treatment to protect their feelings here.
As for users leaving, this happens all the time: https://lobste.rs/moderations?moderator=(All)&what[users]=users.
In regards to Valsorda, I have already speculated his departure could be due to other factors on this site.
score 3 · canonical
I expected this line of retort, but no, you are wrong.
The fact that the “chilling effect” was originally described in application to sovereign nation law does not matter one bit with respect to the fact that the same effect may also happen under different circumstances, in a different kind of society or community, on a different scale.
Lobste.rs does not need to have a law backed by violence for its community to be susceptible to the chilling effect on discourse.
This sort of reductionist attitude (“well, it’s just how it is, nobody owes you anything, deal with it”) is neither valid nor helpful. Clearly, on Lobste.rs, some feelings are entitled to protection, and some matters deserve special treatment, otherwise our “rules” page would have been empty.
As a community, Lobste.rs can self-regulate and can also be externally regulated by moderators to foster a better community or steer the discourse in a certain way. Disallowing or curbing the abuse of specific ways of expression (like inflammatory, ill-defined tags) to avoid splintering the community and draining it of valuable contributions/participants feels like a valid topic on which the community should (self-)regulate.
score 6 · canonical