We don't have a similar rule yet about article content but my sense is that the community mostly doesn't want to read it—or, to put it a bit more conservatively, sharply discounts it. This is why we see so many "just show me the prompt" responses, along with others like this: https://news.ycombinator.com/genai-pushback. I built that list so I have something to send to users who email about why their genai articles got flagged.
A fascinating arms race is happening: the AIs are training on the humans but the human hivemind is also training on the AIs. Readers are developing allergic sensitivities to language that sounds like an LLM produced it. The AIs will adapt to this, but the humans will adapt in turn. Where it ends up is anyone's guess. I have an optimistic view, but I've already been wrong about this so many times that I have low confidence in it.
The near-term picture is that there is a class distinction between writing (and writers) that use genai vs. writing that does not. That is, as soon as the allergic "this sounds like an LLM" reaction kicks in, the writing immediately gets relegated to a low-status bucket in the reader's mind. That doesn't mean it won't still get looked at - but it is now under a stigma.
(I was rather pleased with the originality of this observation until I remembered pg had written about "writes and write-nots" in https://paulgraham.com/writes.html. Oh well, it's the point that matters.)
This has the happy flipside that anyone who would like readers to classify their article in a high-status bucket, rather than a low-status one, can apply the judo move of simply writing it themselves.
None of what I'm saying is a dismissal of LLM technology per se. We rely on it heavily, and there's no question that it's useful. The question is how best to use it (pg again: https://x.com/paulg/status/2058871512451412457) and whether one should use it to generate or edit (<-- that bit is important) writing which one publishes to human readers.
To turn to OP's questions:
> Should HN add the ability to flag articles as AI-generated? [...] it could just show up as an indicator
Flagging as "just an indicator" would be tagging, which we've always resisted adding to HN, but I wouldn't rule it out.
I hope HN doesn't get into moderating the politics of articles.
I can see a grim future (present?) where "AI generated" turns into a slur, warranted or otherwise, in a world where the difference between human trained to talk like an AI and AI masquerading as human becomes increasingly difficult to discern, and some hidden cabal passes judgement.
That is wholly different from taking a stance on HN being a place for humans to comment on articles.
The times I've pointed point out pretty blatant AI comments, I get nuked with downvotes. So often that I've stopped pointing them out.
Before you say I'm just falsely calling them out, it's typical ChatGPT style of either very amicable or Nobel Laureate tone, lots of formatting, with a couple of paragraphs and then a clever one-line punchline at the end. If you look at those commenters their history, it's all like that. Either generated or assisted. For older accounts you can see the steep increase of it around 2025ish.
Seems like HN absolutely adores AI comments and the rule banning them is (sadly) unnecessary. Or at least not what 'the people' want.
Out of curiosity I looked through your post history to find an example of a time you got downvoted for calling out AI comments. The first one I could find was 3 months ago (https://news.ycombinator.com/item?id=47493096), where you got downvoted for calling out an AI comment...to a comment that had zero common signs of AI writing.
The OP then replied:
> Not AI. Not sure how I feel getting my writing style called out like that though :D
IMO posting "This article is AI" does not add anything to the conversation.
The HN guidelines[1] include:
Please don't complain about tangential annoyances—e.g. article or website formats, name collisions, or back-button breakage.
and
Please don't post insinuations about astroturfing, shilling, brigading, foreign agents, and the like. It degrades discussion and is usually mistaken.
I'd argue pointing out that you think an article is AI is very similar in value to pointing out any of the above. None of us like AI slop. But I wouldn't be surprised if, by the end of 2026, 90-95% of articles posted online are AI slop. Pointing it out is useless. As useless as pointing out that the article breaks the scrollbar (which happens often) or that the article is formatted badly or has poor text contrast, or that an article is Chinese propaganda. Probably true, but posting about it adds nothing to the discussion, and is not allowed on HN.
All we really need is to add "Don't complain that an article is AI" to the guidelines.
Regarding 1, I think a) a sizeable fraction of voters are not able to recognize AI-generated text b) many who notice don't care, or are willing to overlook it if the premise is interesting enough. (The latter is true for me, on occasion)
Maybe we need a two-dimensional voting system: good/bad, ai/human. I think the second axis could cut down on meta-discussions over how much of the article was AI-generated.
Different people weight the slop factor differently, which is the main source of pain at the moment. (For example, another top-level comment in this thread suggests banning AI-generated content outright)
Take any existing article you find insightful and then ask an LLM to "rewrite it, but make better and more engaging". The result is likely "good + slop", assuming the actual content is preserved.
If what people here say is right, and AI usage online is growing and accelerating, then by the end of the year, the vast, vast majority articles posted to the internet (and linked by HN readers) are going to be AI slop, at which point it is not going to matter. "This article is AI" is going to add about as much insight to the conversation as "The author used a spell checker."
> a) a sizeable fraction of voters are not able to recognize AI-generated text b) many who notice don't care, or are willing to overlook it if the premise is interesting enough.
agreed.
For me, the issue I have is that a vocal group seems to despise AI-edited content and they can't manage to take their disgust eleswhere.
Considering YC invests in AI I doubt you’ll get anything of the sort. Too many people here also think you just have to give in and accept (abuser mentality IMO).
I'd be curious if you'd read https://news.ycombinator.com/item?id=48887149 and let me know (once I've finished editing the damn thing) how this changes, or doesn't change, your response.
Programmers have benefited greatly from asymmetrical power structure before AI
"Submit to my knowledge or else!" is abusive.
People can drive a car without needing an expert copilot. Why should they need a software engineer to use a computer?
Spare the appeals to history as the historical record would show software engineers have unemployed many others. Technology moves on; rotary phone makers and travel agents have a seat for in their support group.
Your self selection and vanity could not be more obvious.
I don't think that's a very good comparison. People can drive a car without an expert but they can't build a car. People can use software without being an expert but still require an expert to build software.
AI could provided assistance with building both cars and software, but you still need a decent level of knowledge in either field to get a good result.
Eg: my father, born 1935, still alive, has built double axle trailers, cars and vehicles that work, medical caravans for St Johns Ambulance, ...
His education was six years of primary school and these were all "side projects" for himself or for his community.
> you still need a decent level of knowledge in either field
and the barrier to picking up sufficient knowledge to safely create software, cars, even ground effect flying boats isn't that high for normal human beings .. although it does require a certain can do attitude.
This is something that works better on paper in practice. Namely, there are a hell of a lot of false positives of AI use which frequently causes shitstorms on social media where someone says "AI?" in bad faith and now the OP has to defend themselves and in the case of writing a blog post there aren't as concrete ways to defend yourself. (no, demanding the edit history of the post is not reasonable)
Hacker News adopting such a feature would likely do more harm than good.
Just from looking at the sites in the BlockOrigin-HUGE-AI-Blocklist blocklist, it blocks anything tangent to AI such as Hugging Face and other model developers like Mistral/Z.ai, not sites that actually post AI slop.
That's a good example of the exact problem with such a broad stroke rule.
In general, each blacklist targets a different facet of slop content farm structures. Some LLM use-cases like search are legitimate, but most traffic volume is spam/slop related.
Depends on your use-case, but most people don't need content used in network nuisances like YC AstroTurf posts. =3
Nobody wants to label their stuff as AI generated because they removed credibility. Communities can flag posts as AI generated based on speculation and telltales but it won’t be 100% and will take extra work.
I think the era of the blog is simply dead now and that’s mostly ok. Blogspam and corporate blogs had killed quality bogs ages ago even before AI was a thing. The real question is what replaces it.
Oh and of course the $64k question is this: if an AI generated article is indistinguishable from a human written article and it is accurate and interesting, do you care who wrote it? We want to avoid low quality, not AI generation, right?
Even if you did, how would you even enforce it? Say it was a pure text article, do you count the number of em dashes? Even AI detection scanners purpose built for this are extremely faulty.
I'm of the deepest conviction AI-generated text should not show up at all. Proving that however can be difficult (obvious LLM tells aside). Requiring evidence of authentic human authorship is also difficult, though increasingly I lean towards communities where that is a given for any legitimate shares.
The voting system could be enough if downvoting was added.
AI writing is not the problem - low effort is the problem. Low effort AI articles are full of tics which are obvious, if you've done a lot of AI writing. To write well with AI you need to spend a good deal of time editing.
If you submit something that's low effort but has a clickbait headline that appeals to HN, you may well make the front page even if the article is lightweight (it does happen!) This is true both for AI and for human written articles.
On the flip side, somebody could spend an enormous amount of effort creating a masterpiece with AI. Penalizing that because of the tool that was used is arbitrary.
A problem I see is that what someone may consider to be AI-generated actually isn't. And the AI checkers aren't reliable enough to definitely enough say something is AI-generated.
Yup, and calling out a human-written article as AI-generated would be a serious insult. AI-flagging would incur bigger damage to the community than just having AI-generated contents around.
If something like this was implemented, the benefit of the doubt would have to be given in ambiguous cases, but I don't think it's that hard to tell most of the time.
The latest codemaxxed models all tend to write in very distinct, instantly recognizable ways unless carefully instructed otherwise (honestly a good thing if you want to avoid wasting time reading AI text). A great example is this submission that's currently #1 on the front page (which is also just a thinly veiled advertisement): https://news.ycombinator.com/item?id=48884853
The regular voting system is not enough because posts can't be downvoted and for some reason many people are not bothered by the notion of reading something no one bothered to write.
The issue is complicated by the fact that there can be substantial effort invested in a process outside of the writing itself - and so AI written does not guarantee that the content will not valuable. But I'm inclined to punish it anyway to establish a norm of valuing genuine human communication. I think this norm has always been present but we didn't know until we'd really explored the alternatives.
I spend a LOT of time reading AI generated content because I use AI a lot for various purposes - maybe I'm more sensitive to its voice than some. AI voice always bothers me and its been getting more annoying the more I notice it, but there is a huge difference in reading responses to my own prompts and in reading the response to a prompt I haven't seen, when I don't know how many revisions there were, when I don't know if a human mind reviewed it at all before clicking send.
It becomes an unacceptable distraction because I don't know if I'm investing more time in the content than the author did, when in normal written communication the author would be putting in at least 5x the work.
I get annoyed when I see AI telltale signs too, but for example in my case I type 10x more than the final piece, is it really AI generated or just reworded? I don't use AI to fix my comments, just when I want to format article size pieces I post on my blog.
People keep asking this but I don't see how this is even a consideration - it isn't going to have AI voice if AI didn't write it. To some extent humans have picked up some of these tells but part of the writing process involves reading your own work, noticing things that are awkward, and rewriting them. If you aren't investing in the writing process even that much, I don't really want to read it anyway.
My entire position: I'm not interested in reading text that sounds like it was written by AI.
As far as I understand, "flagging" is intended for things that break the guidelines. AI-generated content is explicitly banned, but only in comments, not submissions.
I think it would be great if AI-generated submissions were outright banned as they fundamentally break the balance of effort that HN was built upon, but as was already stated by another user, YC is heavily invested in AI so there's a conflict of interest there.
Not posting for self promotion is in the guidelines. I have to assume anyone pushing out AI articles is primarily looking for traffic and self promotion, as they didn’t care enough about the topic to write their own opinion.
Self promotion isn't banned, otherwise Show HNs wouldn't exist. The exact guideline, emphasis mine:
> Please don't use HN primarily for promotion. It's ok to post your own stuff part of the time, but the primary use of the site should be for curiosity.
Do the guidelines even matter if they're not enforced? This place is drowning in AI shovelware submissions from accounts that are only used for self promotion.
This could be a false memory, but I think it’s the idea that the submission downvotes should be reserved for things that break the rules, and in that case, the post should be flagged instead.
The recent rule addition to the Guidelines says this: "Don't post generated text or AI-edited text. HN is for conversation between humans." And I think that covers comments, but I'd be happy to see it also cover articles that are blatantly and primarily if not exclusively AI-generated. But how much AI is permitted? For instance: I'm writing a blog post now. It's all mine. If I include an AI-generated cartoon at the end, just to illustrate something, but not to be the whole or primary point of the article, is that AI-generated? Would the rule be conservative in nature to the extent that mostly human but clearly also AI-enhanced might get flagged but it's in the discretion of the moderators? How would you propose enforcing as to articles (versus comments which are usually quite obvious and thankfully have pretty much stopped being AI-generated since the rule was implemented, for the most part)?
Most parsimonious explanation IMV: site staff can't see most AI slop. Reasons unimportant, but moderation systems are guaranteed to break down when the moderators themselves have poor classification ability.
A simple beneficial step that would lead to modest improvements and little downside: partner with Pangram. Either adding it as an automated spam filter, or by simply attaching the detection % to all posts.
This makes sense if AI articles are bad or low quality, but what if one day, the AI generated content is actually good? As good or even better than what any human creates?
Is it purely just a "human supremacist" desire that fuels the motivation to ban or block such articles?
That will only further increase the stigma surrounding LLMs. On Lobsters it actually got to the point where I no longer felt welcome on the site, even though I don't use LLMs to generate articles. The constant "this is AI slop" commentary is noisy and tiresome as well.
Saw three years worth of my work be classified as "slop" because a few months ago I subscribed to and started using Claude. Got flagged so much I became the second most flagged user on the site according to their own statistics. Got told to delete my account by the website itself. Got directly told to leave by other users. I think the only reason I didn't is enough people messaged me privately to ask me to stay.
It got to the point I'm sitting on reviewed and tested patches for Mesa that I'm too ashamed to submit because of Claude's participation.
Humans? We're not particularly effective at this as a whole...
AI service ? We'd probably have to pay for that AI to detect that AI and well.. Its also not particularly effective
Effectiveness is important, because we dont want real human produced data to be accidentally removed from view, just as much if not more so than having AI gen data being left on the site.
Setting aside that this is subjective, I think it’s safe to say that from the POV of most of this site’s target audiences, that “start” happened a long time ago. PG’s essay on submarine articles didn’t come out of nowhere, and he hasn’t been active here in…a decade? Ish?
We don't allow genai text on HN itself - see https://news.ycombinator.com/newsguidelines.html#generated and https://news.ycombinator.com/item?id=47340079. How to enforce this is a separate question, of course, but the rule exists.
We don't have a similar rule yet about article content but my sense is that the community mostly doesn't want to read it—or, to put it a bit more conservatively, sharply discounts it. This is why we see so many "just show me the prompt" responses, along with others like this: https://news.ycombinator.com/genai-pushback. I built that list so I have something to send to users who email about why their genai articles got flagged.
A fascinating arms race is happening: the AIs are training on the humans but the human hivemind is also training on the AIs. Readers are developing allergic sensitivities to language that sounds like an LLM produced it. The AIs will adapt to this, but the humans will adapt in turn. Where it ends up is anyone's guess. I have an optimistic view, but I've already been wrong about this so many times that I have low confidence in it.
The near-term picture is that there is a class distinction between writing (and writers) that use genai vs. writing that does not. That is, as soon as the allergic "this sounds like an LLM" reaction kicks in, the writing immediately gets relegated to a low-status bucket in the reader's mind. That doesn't mean it won't still get looked at - but it is now under a stigma.
(I was rather pleased with the originality of this observation until I remembered pg had written about "writes and write-nots" in https://paulgraham.com/writes.html. Oh well, it's the point that matters.)
This has the happy flipside that anyone who would like readers to classify their article in a high-status bucket, rather than a low-status one, can apply the judo move of simply writing it themselves.
None of what I'm saying is a dismissal of LLM technology per se. We rely on it heavily, and there's no question that it's useful. The question is how best to use it (pg again: https://x.com/paulg/status/2058871512451412457) and whether one should use it to generate or edit (<-- that bit is important) writing which one publishes to human readers.
To turn to OP's questions:
> Should HN add the ability to flag articles as AI-generated? [...] it could just show up as an indicator
Flagging as "just an indicator" would be tagging, which we've always resisted adding to HN, but I wouldn't rule it out.
> Why is the regular voting system not enough?
The regular voting system is never enough. (https://hn.algolia.com/?dateRange=all&page=0&prefix=false&so...)
[editing - bear with me... [you guys are adding replies faster than I can finish my own damn comment...]]
I hope HN doesn't get into moderating the politics of articles.
I can see a grim future (present?) where "AI generated" turns into a slur, warranted or otherwise, in a world where the difference between human trained to talk like an AI and AI masquerading as human becomes increasingly difficult to discern, and some hidden cabal passes judgement.
That is wholly different from taking a stance on HN being a place for humans to comment on articles.
You've accurately described the current state of Lobsters.
The times I've pointed point out pretty blatant AI comments, I get nuked with downvotes. So often that I've stopped pointing them out.
Before you say I'm just falsely calling them out, it's typical ChatGPT style of either very amicable or Nobel Laureate tone, lots of formatting, with a couple of paragraphs and then a clever one-line punchline at the end. If you look at those commenters their history, it's all like that. Either generated or assisted. For older accounts you can see the steep increase of it around 2025ish.
Seems like HN absolutely adores AI comments and the rule banning them is (sadly) unnecessary. Or at least not what 'the people' want.
Out of curiosity I looked through your post history to find an example of a time you got downvoted for calling out AI comments. The first one I could find was 3 months ago (https://news.ycombinator.com/item?id=47493096), where you got downvoted for calling out an AI comment...to a comment that had zero common signs of AI writing.
The OP then replied:
> Not AI. Not sure how I feel getting my writing style called out like that though :D
IMO posting "This article is AI" does not add anything to the conversation.
The HN guidelines[1] include:
and I'd argue pointing out that you think an article is AI is very similar in value to pointing out any of the above. None of us like AI slop. But I wouldn't be surprised if, by the end of 2026, 90-95% of articles posted online are AI slop. Pointing it out is useless. As useless as pointing out that the article breaks the scrollbar (which happens often) or that the article is formatted badly or has poor text contrast, or that an article is Chinese propaganda. Probably true, but posting about it adds nothing to the discussion, and is not allowed on HN.All we really need is to add "Don't complain that an article is AI" to the guidelines.
1: https://news.ycombinator.com/newsguidelines.html
Regarding 1, I think a) a sizeable fraction of voters are not able to recognize AI-generated text b) many who notice don't care, or are willing to overlook it if the premise is interesting enough. (The latter is true for me, on occasion)
Maybe we need a two-dimensional voting system: good/bad, ai/human. I think the second axis could cut down on meta-discussions over how much of the article was AI-generated.
Why do we need anything more than good/bad?
If there is a great post on a topic and the author used AI when generating it, what’s so bad about that?
Different people weight the slop factor differently, which is the main source of pain at the moment. (For example, another top-level comment in this thread suggests banning AI-generated content outright)
What’s different between “slop” and “bad”
Take any existing article you find insightful and then ask an LLM to "rewrite it, but make better and more engaging". The result is likely "good + slop", assuming the actual content is preserved.
> many who notice don't care, or are willing to overlook it if the premise is interesting enough
I imagine the set of articles that are somehow both interesting-enough-to-read but not interesting-enough-to-write is smaller than you'd think.
In most cases the bar is not "is it worth reading" but "is it worth discussing"
If what people here say is right, and AI usage online is growing and accelerating, then by the end of the year, the vast, vast majority articles posted to the internet (and linked by HN readers) are going to be AI slop, at which point it is not going to matter. "This article is AI" is going to add about as much insight to the conversation as "The author used a spell checker."
> a) a sizeable fraction of voters are not able to recognize AI-generated text b) many who notice don't care, or are willing to overlook it if the premise is interesting enough.
agreed.
For me, the issue I have is that a vocal group seems to despise AI-edited content and they can't manage to take their disgust eleswhere.
AI-editing is another tool, just like spellcheck.
Considering YC invests in AI I doubt you’ll get anything of the sort. Too many people here also think you just have to give in and accept (abuser mentality IMO).
I'd be curious if you'd read https://news.ycombinator.com/item?id=48887149 and let me know (once I've finished editing the damn thing) how this changes, or doesn't change, your response.
Programmers have benefited greatly from asymmetrical power structure before AI
"Submit to my knowledge or else!" is abusive.
People can drive a car without needing an expert copilot. Why should they need a software engineer to use a computer?
Spare the appeals to history as the historical record would show software engineers have unemployed many others. Technology moves on; rotary phone makers and travel agents have a seat for in their support group.
Your self selection and vanity could not be more obvious.
I don't think that's a very good comparison. People can drive a car without an expert but they can't build a car. People can use software without being an expert but still require an expert to build software.
AI could provided assistance with building both cars and software, but you still need a decent level of knowledge in either field to get a good result.
> People can drive a car without an expert
Not every person can drive a car.
> but they can't build a car.
Not every person is unable to build a car.
Eg: my father, born 1935, still alive, has built double axle trailers, cars and vehicles that work, medical caravans for St Johns Ambulance, ...
His education was six years of primary school and these were all "side projects" for himself or for his community.
> you still need a decent level of knowledge in either field
and the barrier to picking up sufficient knowledge to safely create software, cars, even ground effect flying boats isn't that high for normal human beings .. although it does require a certain can do attitude.
This is something that works better on paper in practice. Namely, there are a hell of a lot of false positives of AI use which frequently causes shitstorms on social media where someone says "AI?" in bad faith and now the OP has to defend themselves and in the case of writing a blog post there aren't as concrete ways to defend yourself. (no, demanding the edit history of the post is not reasonable)
Hacker News adopting such a feature would likely do more harm than good.
uBlock Origin users already filter a large portion of "AI" slop sites. Saves a lot of time when searching for something specific. =3
https://github.com/alvi-se/ai-ublock-blacklist
https://github.com/laylavish/uBlockOrigin-HUGE-AI-Blocklist
https://github.com/Stevoisiak/Stevos-AI-Blocklist/
Just from looking at the sites in the BlockOrigin-HUGE-AI-Blocklist blocklist, it blocks anything tangent to AI such as Hugging Face and other model developers like Mistral/Z.ai, not sites that actually post AI slop.
That's a good example of the exact problem with such a broad stroke rule.
In general, each blacklist targets a different facet of slop content farm structures. Some LLM use-cases like search are legitimate, but most traffic volume is spam/slop related.
Depends on your use-case, but most people don't need content used in network nuisances like YC AstroTurf posts. =3
Being carelessly overzealous with flags/blocking is a reliable way to kill a community by making it overly toxic, speaking from experience.
Nobody wants to label their stuff as AI generated because they removed credibility. Communities can flag posts as AI generated based on speculation and telltales but it won’t be 100% and will take extra work.
I think the era of the blog is simply dead now and that’s mostly ok. Blogspam and corporate blogs had killed quality bogs ages ago even before AI was a thing. The real question is what replaces it.
Oh and of course the $64k question is this: if an AI generated article is indistinguishable from a human written article and it is accurate and interesting, do you care who wrote it? We want to avoid low quality, not AI generation, right?
Even if you did, how would you even enforce it? Say it was a pure text article, do you count the number of em dashes? Even AI detection scanners purpose built for this are extremely faulty.
I'm of the deepest conviction AI-generated text should not show up at all. Proving that however can be difficult (obvious LLM tells aside). Requiring evidence of authentic human authorship is also difficult, though increasingly I lean towards communities where that is a given for any legitimate shares.
What qualifies as AI generated? If a human writes it then has AI improve/fix it, does that count?
How do you tell which is the case?
If we don't allow AI help at all, is that perhaps discriminating against those who don't feel comfortable posting with imperfect English?
I agree in principle, but am concerned in implementation... I'm not sure we can be fair without high risk of discrimination
Edit: typo fix
Edit: or am I AI?! And making edits looks more legit.... (To be clear: I'm not, I play by rules)
>increasingly I lean towards communities where that is a given for any legitimate shares.
I have a hard time finding these communities
Mastodon, but the hard part is discovery for sure.
The voting system could be enough if downvoting was added.
AI writing is not the problem - low effort is the problem. Low effort AI articles are full of tics which are obvious, if you've done a lot of AI writing. To write well with AI you need to spend a good deal of time editing.
If you submit something that's low effort but has a clickbait headline that appeals to HN, you may well make the front page even if the article is lightweight (it does happen!) This is true both for AI and for human written articles.
On the flip side, somebody could spend an enormous amount of effort creating a masterpiece with AI. Penalizing that because of the tool that was used is arbitrary.
Might be more appropriate to add a "not AI" flag at this point.
A problem I see is that what someone may consider to be AI-generated actually isn't. And the AI checkers aren't reliable enough to definitely enough say something is AI-generated.
Yup, and calling out a human-written article as AI-generated would be a serious insult. AI-flagging would incur bigger damage to the community than just having AI-generated contents around.
If something like this was implemented, the benefit of the doubt would have to be given in ambiguous cases, but I don't think it's that hard to tell most of the time.
The latest codemaxxed models all tend to write in very distinct, instantly recognizable ways unless carefully instructed otherwise (honestly a good thing if you want to avoid wasting time reading AI text). A great example is this submission that's currently #1 on the front page (which is also just a thinly veiled advertisement): https://news.ycombinator.com/item?id=48884853
The regular voting system is not enough because posts can't be downvoted and for some reason many people are not bothered by the notion of reading something no one bothered to write.
The issue is complicated by the fact that there can be substantial effort invested in a process outside of the writing itself - and so AI written does not guarantee that the content will not valuable. But I'm inclined to punish it anyway to establish a norm of valuing genuine human communication. I think this norm has always been present but we didn't know until we'd really explored the alternatives.
I spend a LOT of time reading AI generated content because I use AI a lot for various purposes - maybe I'm more sensitive to its voice than some. AI voice always bothers me and its been getting more annoying the more I notice it, but there is a huge difference in reading responses to my own prompts and in reading the response to a prompt I haven't seen, when I don't know how many revisions there were, when I don't know if a human mind reviewed it at all before clicking send.
It becomes an unacceptable distraction because I don't know if I'm investing more time in the content than the author did, when in normal written communication the author would be putting in at least 5x the work.
I get annoyed when I see AI telltale signs too, but for example in my case I type 10x more than the final piece, is it really AI generated or just reworded? I don't use AI to fix my comments, just when I want to format article size pieces I post on my blog.
People keep asking this but I don't see how this is even a consideration - it isn't going to have AI voice if AI didn't write it. To some extent humans have picked up some of these tells but part of the writing process involves reading your own work, noticing things that are awkward, and rewriting them. If you aren't investing in the writing process even that much, I don't really want to read it anyway.
My entire position: I'm not interested in reading text that sounds like it was written by AI.
Maybe just adding down-vote to submissions would do?
We have “flag”
As far as I understand, "flagging" is intended for things that break the guidelines. AI-generated content is explicitly banned, but only in comments, not submissions.
I think it would be great if AI-generated submissions were outright banned as they fundamentally break the balance of effort that HN was built upon, but as was already stated by another user, YC is heavily invested in AI so there's a conflict of interest there.
Not posting for self promotion is in the guidelines. I have to assume anyone pushing out AI articles is primarily looking for traffic and self promotion, as they didn’t care enough about the topic to write their own opinion.
Self promotion isn't banned, otherwise Show HNs wouldn't exist. The exact guideline, emphasis mine:
> Please don't use HN primarily for promotion. It's ok to post your own stuff part of the time, but the primary use of the site should be for curiosity.
Do the guidelines even matter if they're not enforced? This place is drowning in AI shovelware submissions from accounts that are only used for self promotion.
Hn rarely allows people to downvote,only after you become an active member for many years
And im sure this was designed in order to encourage positive discussion
As for the ability to downvote comments, my understanding is that the only requirement is having at least 500 karma.
There is no longer a downvote on submissions.
Do you know the reason? Allowing down voting on comments but not on submissions?
This could be a false memory, but I think it’s the idea that the submission downvotes should be reserved for things that break the rules, and in that case, the post should be flagged instead.
The recent rule addition to the Guidelines says this: "Don't post generated text or AI-edited text. HN is for conversation between humans." And I think that covers comments, but I'd be happy to see it also cover articles that are blatantly and primarily if not exclusively AI-generated. But how much AI is permitted? For instance: I'm writing a blog post now. It's all mine. If I include an AI-generated cartoon at the end, just to illustrate something, but not to be the whole or primary point of the article, is that AI-generated? Would the rule be conservative in nature to the extent that mostly human but clearly also AI-enhanced might get flagged but it's in the discretion of the moderators? How would you propose enforcing as to articles (versus comments which are usually quite obvious and thankfully have pretty much stopped being AI-generated since the rule was implemented, for the most part)?
How about if you did part of your research with AI, then typed the text yourself?
Asking questions like that nearly got me kicked out of Lobsters.
Most parsimonious explanation IMV: site staff can't see most AI slop. Reasons unimportant, but moderation systems are guaranteed to break down when the moderators themselves have poor classification ability.
A simple beneficial step that would lead to modest improvements and little downside: partner with Pangram. Either adding it as an automated spam filter, or by simply attaching the detection % to all posts.
This makes sense if AI articles are bad or low quality, but what if one day, the AI generated content is actually good? As good or even better than what any human creates?
Is it purely just a "human supremacist" desire that fuels the motivation to ban or block such articles?
That will only further increase the stigma surrounding LLMs. On Lobsters it actually got to the point where I no longer felt welcome on the site, even though I don't use LLMs to generate articles. The constant "this is AI slop" commentary is noisy and tiresome as well.
Why would you no longer feel welcome?
Saw three years worth of my work be classified as "slop" because a few months ago I subscribed to and started using Claude. Got flagged so much I became the second most flagged user on the site according to their own statistics. Got told to delete my account by the website itself. Got directly told to leave by other users. I think the only reason I didn't is enough people messaged me privately to ask me to stay.
It got to the point I'm sitting on reviewed and tested patches for Mesa that I'm too ashamed to submit because of Claude's participation.
There absolutely needs to be stigma surrounding LLM generated work that is masquerading as creativity.
AI slop is AI slop.
Great so I can use a CSS rule to hide anything with the flag.
Sounds like a good job for AI. Why should humans have to waste their time on it? Accounts that post any should just get banned and deleted.
how are we detecting AI gen text?
Humans? We're not particularly effective at this as a whole...
AI service ? We'd probably have to pay for that AI to detect that AI and well.. Its also not particularly effective
Effectiveness is important, because we dont want real human produced data to be accidentally removed from view, just as much if not more so than having AI gen data being left on the site.
+1, I would love to stop reading AI slop.
> why is the regular voting system not enough
Voting systems can be gamed and as HN becomes bigger and bigger it'll start to attract unsavory audiences who have an agenda.
Setting aside that this is subjective, I think it’s safe to say that from the POV of most of this site’s target audiences, that “start” happened a long time ago. PG’s essay on submarine articles didn’t come out of nowhere, and he hasn’t been active here in…a decade? Ish?
My original account dated to 2010 and even then he wasn't very active
Similar discussion on the other site: https://lobste.rs/s/ktew3s/who_does_anubis_actually_stop#c_c...