Open questions:
1. Why is the regular voting system not enough?
2. Should HN change in response to the gen AI era? It has been successful not changing fundamentals.
Open questions:
1. Why is the regular voting system not enough?
2. Should HN change in response to the gen AI era? It has been successful not changing fundamentals.
83 comments
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 more conservatively, 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.
It's a fascinating arms race right now: 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.
For the present, there is an emerging class distinction between writing (and writers) that use genai vs. writing that does not. As soon as the "this sounds like an LLM" allergy kicks in, the writing instantly 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 until I remembered pg had come up with "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 as high-status rather than low-status can apply the judo move of simply writing it themselves.
Now I need to add the disclaimer that none of this 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 to use it (pg again: https://x.com/paulg/status/2058871512451412457) and whether one should use it on writing that one publishes to other humans.
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.
What I do think we'll (finally) add is a "please give a reason why you flagged this post" step, and "because I think it's genai" will be one choice among several (spam, offtopic, mean, etc.)
> 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...
> Should HN change in response to the gen AI era?
To this I am tempted to reply with https://news.ycombinator.com/item?id=48887149 in homage to https://news.ycombinator.com/item?id=3742902.
Hackernews isn’t work, obviously, but “it’s impossible to engage deeper with the material because the author doesn’t really exist” is sort of a problem for a discussion site. If the human coauthor puts in enough work, they can make sure the doc really reflects their views and their understanding, but in my experience that’s much less common.
You can't have a discussion about it, how it was done, because for 80% of cases it's "the AI did it".
I don't think this is that interesting or useful for HN, because indeed it's a discussion site.
It would be different if the vibe coded thing didn't come as a "look what it made and its vibe written README" but instead as a human written blog article about "how I made this thing using AI, what models, prompts and harnesses, how the experience was etc etc etc" --- again that would be something interesting to have a discussion about.
But otherwise what is there to say except "yeah cool that's cool that the AI made you that thing"
> it’s impossible to know whether misconceptions or gaps are because the author doesn’t understand the material deeply enough or the author does but the AI doesn’t and the author’s not proofreading carefully enough. Or if a surprising idea is raised — is it the authors insight, can they elaborate on it, where did it come from, etc?
Ways to contribute: Peer review, human code review, money for tokens for code and code review, configure SAST and DAST tools to make it DevSecOps instead of DevOpsSec, configure AGENTS.md and an /upgrade-dependencies-and-run-tests skill,
Compare "It's inadequate because it's AI" with "it looks like this generation of agents doesn't yet handle this quality aspect with or without explicit prompts" with "you could improve quality by writing tests, docs, and before that refactor for maintainability and subjective elegance" (if you're trying develop a hit open source project)
> impossible to know whether misconceptions or gaps are because the author doesn’t understand the material
Shouldn't a researcher consider the argument - it's premises and form - instead of tangential ad hominem about the author?
Pushing for well-formed arguments is older than AI
“The author doesn’t understand the material” isn’t an ad hominem — most of my work involves working with other people, usually less experienced, to help them deliver reasonably complex econometric or ML models. Sometimes I’m hands on, usually not. Knowing if it’s “the author not understanding the material” is important because my job is to help them understand it so that they can deliver something that 1) works and 2) does what it’s supposed to. AI written docs are at best a mixed blessing.
Or maybe it’s me not understanding! Either way, if something looks wrong, I want to discuss it with people who know the subject matter and can talk me through the parts that look off so that I can learn from it.
Thanks for giving shape to a general annoyance, in my case with code, that I only recently started noticing.
When discussing code with my colleagues, especially with those that I do not know particularly well, I often relied on the quality of the code they produced to modulate how technical the conversation ought to be be.
Now, instead, I often see very complex code from people that I know wouldn't have been able to produce it themselves, and I have no clue how to engage with it, how to review it with them, how detailed can my comments be etc.
It's pretty annoying.
FWIW, same problem with PRs or PoC that I have to work on; now my first question is, "did you know about his behaviour?". The first step, getting a decent spec, is delayed to after a first draft implementation is already pushed...
It doesn’t matter. The human responsible for the written piece is … responsible for the written piece. If someone is willing to use AI to supplement their own intellect and/or writing abilities, then they need to take ownership of the mistakes, misconceptions, etc., that come along with that, the same as an author who does all the work themselves. The problem you’ve identified here is that the human responsible for the writing is not fully taking responsibility for the writing.
Perhaps AI-written articles should include the prompt as hidden text on the page. Or include a link with the prompt embedded, that causes the AI to spit out the same article that you could then ask questions about.
“Why didn’t the author consider [this thing that seems similar]” is almost never something you can ask an llm, especially for articles where the author is putting in so little of their own effort that they don’t edit out obvious AI tells.
It's always been the case that the HN audience has been "allergic" to low effort slop (including human generated).
The OP (and my) point was trying to point out how the AI generated stuff might masquerade as something "interesting" if it had been human written, but actually is much of a nothing burger when there's no human behind it:
if a human made a strange, weird, or even sub-optimal design decision, that's something to possibly have an interesting discussion about.
if an AI does the same, first of, I don't care and don't wanna guess why its weights did the thing it did, it's just not as interesting as human motivations, UNLESS the main meat of the article/link is in fact about (again the human experience of) "how I made this thing, using AI, what prompts, harness, models, etc".
I hope it makes sense and allows you to no longer dismiss this as an "allergy" (aka the audience's problem), but as something that is actually a quality problem with the post.
It is, however, my experience that even the people who spend a lot of time and effort generating something with AI, tend to shrug when you ask a deeper question about the project, and often point to the AI generated documentation, which is, let me be subtle and say, "not really a joy to read", and especially pales in comparison to having a conversation with a human who spent their effort on learning how to do the thing it did, instead of prompting an AI.
I'm also not saying there aren't any exceptions to this. But I've found them rare.
Like I said, even if the human doesn't quite understand what they generated, even then it would be preferable to read a human written article about how it was generated, using what tools and harnesses, etc.
I come to this website to learn stuff and converse with other people, not just to go "ooh aah" at what somebody generated with AI (if that is all they can offer).
Hope that makes sense.
It's not a purity test, it's as the author is communicating they don't care whether the reader has any signals of what is accurate vs inaccurate information, which puts the burden of investigating how much is accurate on the reader at every step when there's some minimum expectation that should be an author's role (outside of topics where there is some expectation of ulterior motives/biases and one would naturally engage more critically minded).
When people complain here it's more often than not when an article has no disclaimer about AI use or what has been human-reviewed, so the burden again falls on the reader who is now even more skeptical. That is more to ask of a reader than when it's coming from say a known expert and the reader is receptive to engage and learn.
That's the reason tired cliches and turns of phrase (overused by LLMs) have become a heuristic for whether to pay attention, because it's a sign that there's some unknown quantity of of the article that hasn't had human review and it's easier to put in the bucket of 'maybe worthwhile but would need a fully human analysis of this' or just outright rejection (as we've seen from comments).
Edit: I see a sibling comment has raised the same observation.
Like, if some non-controversial article makes a statement about something technical (where one's guard isn't already raised) but you've observed signs of LLM use (without any disclosure of to what degree) then instead of thinking it might be an interesting thing to follow-up on or remember one might be thinking instead 'is this something the author themselves understands and has reviewed for accuracy or slipped in by the LLM' and other such distractions (and legwork if wanting to try and fact-check such things on the spot).
It goes from having a perhaps pleasurable, educational read to questioning and being more skeptical/cynical about the material. HN's guidelines meanwhile encourage good faith engagement, which is challenging.
Edit: corrected permalink (had accidental extra digit).
[1] https://news.ycombinator.com/item?id=48888422
Does that mean that an article being AI-generated is a flaggable offense? Should we be flagging suspected AI-generated articles already, or should we wait for the flagging system to support reasons first?
I don't know for sure yet. But you're right that that would follow from what I said.
What I do know is that users should flag articles that they think don't belong on HN. But this should rely on their reading of https://news.ycombinator.com/newsguidelines.html and not just be based on a like/dislike reaction. That is, users should flag articles that they think break the site guidelines, not just ones they dislike.
So either I read your mind, or it is best not to flag unless something is obviously, unequivocally, running afoul of the guidelines.
I keep being annoyed at how arbitrary, and dare I say hypocritical, the entire flagging mechanism and its surrounding politics are.
I've also been aghast at the nerve of those same serial posters applying the letter of the law against other users. I genuinely wouldn't care if Bartosz Ciechanowski only self-promotes every last one of his own posts. Some users are significantly more interesting than others, whether or not they violate the site guidelines marginally on a technicality.
I have suspected it's possible some users flags are worth more than others, and I wouldn't be surprised if my flagging ability was silently removed several years ago.
People love to make vague claims like this based on things we supposedly said. How about you quote what exactly what I literally said so readers can make up their own minds?
Sorry if that is testy (well, it is testy), but I can't count the number of times that disgruntled users have posted falsely to HN threads to vent their residual frustrations with the mods.
> I'm afraid we took vouching privileges away from your account because you vouched for too many comments that were unsubstantive and/or flamebait and/or otherwise broke the site guidelines
> If you want to build up a track record for a while of vouching for good comments only, and then email so we can look at the recent vouches
Wondering how does one tell what is ‘good’ in your mind:
> Btw if you're unsure about a case you can always check with us about it. I know the borderline cases are not always easy to call.
—-
In other words, do as dang himself would do, or get penalised. I am not in the business of mind reading.
This was years ago. I am pretty sure lately I have been flagging/vouching stuff I genuinely believe were OK, though you have all the data in front to smite me with “ah, but on this day you vouched this bad comment! Gotcha!” so in the end it’s always a losing battle.
In any case, my issue is with you saying, I quote, ”users should flag articles that they think don't belong on HN”. Emphasis mine. No. It is more nuanced than that.
I can see how using the word "good" was confusing, but I just meant the opposite of "comments that were unsubstantive and/or flamebait and/or otherwise broke the site guidelines". It just boils down to: what fits the guidelines or not.
> Wondering how does one tell what is ‘good’ in your mind
Since you can't read my mind (at least I assume you can't, and if you could you wouldn't have this question), that's not doable. What you can do is assess things according to your own reading of the site guidelines. If you assimilate https://news.ycombinator.com/newsguidelines.html and evaluate comments based on that and not just on what you like/dislike, then I imagine you won't end up too far from where the mods are, because that's what we're also trying to do.
I'm sure there will always be individual cases where we disagree—even tomhow and I disagree on individual cases—but they should be a minority.
If there are specific cases where you think we got it wrong, I'd be interested to see links, and we're always willing to hear a contrary argument.
Good job being you. Please continue doing so.
I have a lot of respect for his work.
To think that someone has multiple / repeated disagreements with Dang is wild to me.
In my history on this site, I have not once seen Dang say or do anything that’s not in the best interest of the community.
He’s possibly one of the least controversial moderators I have ever encountered online.
I mean it's kinda hard to kick out the VC brainrot get-rich-quick sanity-is-tertiary scum when you're literally housed under a ycombinator subdomain, but that is technically also a choice that people might see as controversial.
It is no coincidence lawyering is a lucrative career, even when (in most countries) it is simply about interpreting what’s written in the big book.
Until you rewrite the guidelines in an unambiguous programming language (Lisp will do), and provide a system to discern whether a comment falls afoul of them, your personal interpretation of the Word will often differ from someone else’s.
That said, thanks for the edifying discussion.
If however you see things isn’t lining up well with how they want things flagged, it makes sense to remove that from you. Unless there’s more punishment attached this seems very sensible regardless of how genuinely you believe in things.
Privately, when I was new he was polite and helpful. (Tried to be polite, was returned.)
/singular anecdote for reader reference
I'm more and more think that link actually hurts the site more than helps. It's a gotcha against everything, can be used a sword and a shield too.
See the part about politics. There is more and more US domestic politics post here (which I actually flag all the time) yet when the topic comes up people will say: "should hackers turn their back on when the world burns" So basically everything goes as long as it fits the narrative.
And the worst offender imo is the last part about reddit. Because I do _really agree_ that HN is more and more like reddit, if not already is. This very post with the comments where everyone is asking for a downvote button is a perfect example of that.
If this is true then this statement should be added to the guidelines. Many a time I've seen a submission and thought "My god, this is the most moronic thing I've ever read in my life, nearly every single factual claim in this is wrong" but I don't flag it because "don't post utter drivel" is not in the guidelines.
It's true that we don't spell things out precisely, but that's because it's impossible to spell things out fully precisely. If you start, where do you stop?
Same as on other sites: Spam and/or bot, abuse and/or insults and other ad hominems and particularly the point in the guidelines where you go: Please don't use Hacker News for political or ideological battle. It tramples curiosity. and Off-Topic: Most stories about politics, or crime, or sports, or celebrities, unless they're evidence of some interesting new phenomenon. If they'd cover it on TV news, it's probably off-topic.
Not that my flagging amounts to anything, there's so much domestic US politics articles from mainstream media on here that I often wonder whether I misunderstand the guidelines, but then, I'm just an Europoor /s
I know you hate the supposedly "noob" comment, as you've put it over the years, and as it is written in the guidelines, that HN is or isn't turning into Reddit, but this aspect of HN makes it indistinguishable to me as a long-time reader here whether you accept it from your perspective as a long-time moderator or not.
Because HN moderation allows people to downvote or flag things simply because they don't like something, it will always be like Reddit. There is no distinguishing feature to separate it otherwise.
One thing that would, would be forcing users to say why they downvoted or flagged something.
Edit: If you spend years saying "no, no, it isn't true" to people, you're just sticking your fingers in your ears. Where there's smoke, there's fire.
- To give an outlet for (dis)agreement: We could have two arrows on each post, one for agree/disagree, the other for 'legitimate post that helps the conversation'. Someone posting wrong info rarely helps the conversation, but also often enough it's an opinion matter and you can disagree while leaving the latter arrow alone. The latter determines sort order (most helpful first); the former is displayed as a fun fact on posts that are deemed helpful (author sees exact counts just like now)
- A system that seems to work better: A local tech site uses a system where you assign a quality level to a post. The short labels are (translated) "flamebait", "irrelevant but well-intentioned", "relevant but well-known", "useful contribution", and "exceptionally useful". Each one has a numerical value, ranging from -1 to +3, and it'll take the median of all votes (with a half-vote bias towards +1 I think, but that's an implementation detail). If your median comes out to -1, it's collapsed by default; at +2 and +3, it gets highlighted. Incorporating opinion into your quality assessment is a problem here, too, but then you might get +0 instead of +1 and not (like on HN) a score of -7. Not just does that feel a lot more fair, the system and its UI also just makes more clear to everyone that you're supposed to objectively score the comments on the mentioned guidelines (it shows a description every time you hover over a vote option)
Said another way: I couldn’t disagree more.
Your comment right now, and mine above are not interesting!
In another thread there might be one whole comment out of hundreds that retells the experience of a senior engineer working on an uncommon problem, sharing specifics about a particular industry, piece of intellectual property, or another individual.
The rest will not be interesting. That's normal! That's OK! But most are not.
You may be more ephemeral with your community identities. That’s cool, too.
Just thinking out loud I suppose.
The fact you can't just make an account, log in and start downvoting is already a massive improvement over Reddit
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.
I am not sure how the future may be, but the direction of AIs has been to solidify their distinct writing style rather than assimilate to humans. So we can start addressing what we are dealing with now. I doubt HN tagging/flagging can put enough pressure to affect how llms will evolve anyway in the future, while the llm speech becomes more and more ubiquitous and unhinged.
Yes, but what if it doesn't? Anyone can create a model or prompt to adjust the tone, remove the tells that people are now on the lookout for, etc.
And it's self-reinforcing, because the people that are on the lookout will always consider themselves to be right in their judgment, also because of survivorship bias - they won't know if they missed an AI generated comment/post unless it's revealed.
This is one reason why I'm in favor of tagging AI generated posts, because at one point we can't tell. If we can't tell, does it still matter? I don't know, but I'd like to know anyway.
Of course, it can't be mandated, so if the submitter doesn't tag it and the readers can't tell the difference, it's futile.
we are not at the point where "we can't tell", and if we actually couldn't tell we wouldn't be complaining about it
The massive 1000+ comments/upvotes AI model release threads only show up on HN. Lobsters doesn't accept "business" articles like that. You're quite likely to find reactionary articles that are critical of AI on the front page though. There are several on the front page right now, in addition to the ones derisively tagged vibecoding.
It'll pass when the next cool tech comes around.
I'm not picking on you - it's practically a universal response, so much so that it must be driven by human hard-wiring. I've written about this so many times that for once I don't even know what to link to. Perhaps https://news.ycombinator.com/item?id=23308098. Or maybe https://news.ycombinator.com/item?id=48427800
The most concerning growing trend I see is comments that don't violate any guidelines, but are flagged to dead.
I'm not talking about posters that are shadowbanned, but about comments that engaged thoughtfully on a topic but the rest of the thread disagreed emphatically with.
I also don't even mean controversial takes on hot-button issues like vaccines.
Just plain old bucking the trend in a thread about AI, transit, housing, layoffs, etc.
I've been browsing with showdead on for as long as I can remember, and I'm seeing this accelerate. Comments from folks both you and I respect as high quality contributors here.
I don't know what the solution is, but it discourages me (and I'm sure others) from having the kind of thought-provoking dialog I've gotten used to reading here for the last 15 years.
Vouch them and/or email to hn@ycombinator.com and say why you think they deserve to be [undead].
I see that a lot, some have been erroneously caught by the real time AI detect filter (which actually isn't too bad at slicing out the actual AI gen comments) others have gone hard against the zeitgeist.
I've had the mods reinstate comments that I've thoroughly disagreed with but were making their best case for an opposing view - the threads are better with the best arguments forward for all sides of the elephant.
I don't know if the vouch threshold has increased or if the mods have removed my ability to vouch for other comments entirely.
Wondering if it was my good faith attempt to inform a user of their comments being auto-flagged [1] before I had any context that they were supposed to be persona non grata.
I can still vouch submissions though
1: https://news.ycombinator.com/item?id=46381180
But yes, it's the "going against the zeitgeist" ones I have in mind. It feels very damaging to the community.
It's a rare individual that reads all the comments .. my impression is they appreciate on point heads up emails about accidentally auto-binned comments and are happy to reverse flagging on good comments, even those that try and make the dark side of a case.
Probably best not to try and flood the zone with anti/pro pet subject matter karen emails though.
FWiW I semi frequently pick apart comments supporting cases I'm in favour of .. because I want to see better arguments being made and to discourage weak easily dismissed shallow support.
Then again, I'm perfectly capable of perceiving bad-faith comments opposing generative AI. I had dismissed the possibility out of hand, so it's possible that the reason I don't see many bad-faith agrees-with-me comments, AI-bashing aside, is because there aren't many: maybe people who agree with me tend to act in good faith.
Vouch, I've perhaps seen reverse things several times (more of a background observation not pursued, so, you know, fuzzy anecdata at best), I did get feedback from @dang that one comment I vouched for was, in fact, a legitimate AI gen comment .. so clearly I'm human enough to fall for the occasional clanker.
(I'm well aware of the genAI policy, and very supportive of it.)
https://news.ycombinator.com/item?id=48660796
Personally that is the future I hope to see. Hence my continuous protestations at the lazy excuses for articles that get posted these days, and the lazier excuses given by their authors for avoiding using their brain.
I was once accused of using AI for writing in the voice of a depressed character because the character had a certain emotional detachment that to the person lodging the accusation indicated AI.
In short it is not just specific phrasing or words but also aesthetic effects that mean one is AI nowadays.
It’s okay to have an opinionated website. Not every corner of the internet needs to be a bastion of free speech
(Subconscious training like when we pick up an accent, though eventually folks might automatically code switch - so that’s hopeful.)
> It's a fascinating arms race right now: 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.
Humans are training to detect AI content. Humans writing more like AIs is an unrelated (and slower) phenomenon.
The class of comments you're talking about—not just ones that say "clanker" literally but the more general category—is particle-wave undecidable right now.
If the commenter is right—i.e. if the commenter they are castigating actually did post LLM-generated text—then it's a community immune system response.
If the commenter is wrong—i.e. they are castigating a sincere human and hauling them to court on false charges—then it's the kind of attack that we tell people not to post here.
They can't know for sure whether what they're saying is true or false, and we can't know for sure how we should moderate it. Both questions depend on information that is unavailable. This is what I mean when I say that the whole question is in a chaotic state right now, and it's too soon to know which way it will stabilize.
I think the decision to ban AI submissions is a good one, but ultimately it's going to create this conflict for some time, maybe a really long time until things stabilize around AI in the broader culture
I hope eventually AI usage does become a taboo, at least in some fields. Creative fields should be for creative humans, not people who can't even publish an article without the help of an LLM
There are tons of ways to use AI that don't intersect with that.
If it becomes very normal and expected that people just let LLMs speak for them online using their voice, then I don't think any kind of online community has a hope of actually keeping that behavior at bay long-term
For people like me that means the end of online communication entirely, most likely. I don't want to talk to people's LLMs
My hope is there's some broader cultural taboos around AI usage for communication purposes at least.
I find LLM-isms to be exactly the same as grammatical errors, but worse. At least when writing before you had to take the effort to type every word, so there was a minimum amount of effort you’d need to expend. If you aren’t catching obvious things like “the honest part” then that likely says bad things about your attention to detail elsewhere.
Stuff interesting enough to get upvoted (i.e. not slop), but I'm so irritated and triggered by pointless, human, comment threads moaning about "this is AI generated".
I really really hope more people take up pen and paper! My last blog post [0] came with proof-of-work attached.
[0] https://abner.page/post/are-we-harold-bloom/
Doesn't seem too hard to fake now, now that AI can generate convincing videos. Failing that, it's definitely within the realm of possibility for AI to create fake pages for a notebook, then generate a blender/unreal engine project to render it.
What a goofy situation to imagine. I hope we can figure out a way though. I personally have no interest in reading anything spat out by an LLM, so if anything can be used to prove that an author wrote something themselves, I'm interested
But nonetheless a problem in desperate need of solving!
Do you believe adding friction to flagging will reduce the quantity of low quality articles?
Or is the flagging of high quality articles a bigger and more pressing problem?
Or is the problem simply too-many-damn-flags?
Just curious.
I guess the idea is that if lots of people flag a comment for the "genai" reason then we can treat that a more precise community signal than "flagged in general". But this argument seems weaker to me as soon as I write it out.
But then again, there’s always reddit :)
Plus, HN's design makes it easy for a single story or kind of story to appear to overwhelm the forum (I say "appear" because really there's more than what's on the frontpage, but most people never bother to check past the frontpage,) which only results in complaints and threads like this. Another layer of organization to handle HN's increasing scale and complexity might be necessary.
Slashdot is still around
That's not a value-judgement of the true content of an article or piece of media, but a fairly objective facet which impacts the HN participant's experience.
It's kind of like how, back in the day, people really wanted to know the filesize of something before they clicked, to avoid a blind-investment of their dial-up bandwidth for indeterminate minutes of waiting (and opportunity-cost of other things not downloaded) that might be more than they really wanted for whatever-it-was.
Then it was to protect our modems bandwidth (cost and time) while now it's more about protecting our own cognitive bandwidth.
Poor writing is not a new thing, of course. Most of the moderation mechanisms that it uses were perfected a quarter century ago when sites like Slashdot were popular as a defense mechanism against bad user behavior. Bad user behavior impacts commenting, article submissions, and moderating itself. While bad users now have AI to abuse, the problem of a large volume of low quality content is is largely the same. The moderation mechanisms end up targeting the problem at the source: identifying good and bad users. So, the same amount of bad users generating a lot more garbage isn't that big of a deal. Getting good karma still is a lot of work and it makes identifying all the garbage created by users without that relatively straightforward.
Using LLMs to tag, flag and filter content might not be a bad thing to experiment with. There are a lot of low quality AI generated opinion pieces that somehow make it to the front page. Same for political and controversial stuff, which of course is against the HN guidelines for content. Auto flagging things that obviously violate guidelines should not be that hard. It's just a matter of having good guard rails. @dang might actually already be doing that. I know I would be if staying on top of piles of generated garbage was part of my job description. It might also be done to give good content a little boost.
The new articles section has a very low signal to noise ratio currently and the window for good content to make it past that is very short. Often articles on the front page will have many duplicate submissions that never made it past that. IMHO duplicate submissions should just count as upvotes on the original. Auto de-duplicating based on canonical URL should not be that hard.
I respectfully disagree on almost all counts (beyond poor writing not being new, of course!).
Moderation mechanisms have not been perfected. They're certainly not perfect, and, given that, I don't know what would make one call them perfected. Humans have probably gotten more accustomed to being moderated, but it'd take a lot to convince me that we have even reached a decent place for moderation at moderate scale, let alone something that is good-to-perfect.
Most importantly, the problem of low quality content is now not the same. Magnitudes matter, and a difference in degree eventually becomes a difference in kind. LLMs have escalated the problem of garbage content beyond what would've previously been conceivable.
To illustrate: How do you dispose of several trash bags at once? Take them to the trash can. How do you dispose of several tens of trash bags at once? Need to rent a dumpster.
Or a classic: If you owe the bank $100, that's your problem. If you owe the bank $100 million, that's the bank's problem.
Bringing it back to written content: doing human-driven moderation on hundreds of submissions a day is tractable with (idk) a couple of people. For thousands or tens of thousands? Intractable. And bear in mind that human-driven moderation is one of the things that keeps HN a better place on the net than many (most) others.
It's definitely not universal. I've seen articles that seem clearly AI-generated, but still get upvoted because the community likes the title/thesis.
I can confirm. Most LLM-written content is low effort, low value. This is somewhat by construction. You get the blandest takes in the blandest language.
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 the HN crowd absolutely adores AI comments and the rule banning them is (sadly) unnecessary. Or at least not what 'the people' want.
The OP then replied:
> Not AI. Not sure how I feel getting my writing style called out like that though :D
So the guidelines are in some sense a red herring.
".. Fault-tolerant and highly available hardware must facilitate low-latency, single-threaded communication with high semantic density in order to achieve multi-dimensional consensus in a safety-critical, heterogeneous, adversarial environment. .."
I am not sure why you think someone saying "not AI trust me bro" carries any merit.
At any rate, like I said, I've given up the war. People enjoy reading that stuff, I'll just be the old man no longer yelling at the clouds.
No LLM has responded like your quote of OP that you assert is LLM generated. Dense verbose writing is not something LLMs do without very specific prompt engineering that no one would do rationally. The tropes of LLM writing are more highly sporadic verbose writing (e.g. "delve") with highly atypical grammatical constructions, it's more precise than "it's sounds like a MBA" like you are asserting. Look at Wikipedia's list and try to figure out which item the OP falls under: https://en.wikipedia.org/wiki/Wikipedia:Signs_of_AI_writing
Yes, OP has a verbose way with words and grammar. On a site like Hacker News that shouldn't be that weird, and accusing people of using AI without a smoking gun hurts culture and discourse far more than doing nothing at all, which is why I'm passionate about baseless AI use accusations. There are more than enough smoking-gun AI-written comments you can flag/downvote respectfully.
I think it's because I'm not optimizing my life to get the correct answer as fast as possible, or to build things as fast as possible.
To me, the most important thing about the internet is connecting with other people. If I ask a question on a forum it's because I want to talk to someone about it, maybe make an acquaintance or even a friend. Otherwise I would of course just ask the AI now. Google has been around for a long time, and could already usually find answers for me. I still would rather discuss with a colleague sometimes than Google every single thing.
Human connection. We need more of it, not less. I think heavy AI use and reliance on AI for thinking, research, communication and building... It's going to isolate people even more
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
Nowadays I usually check the comments first for the "This is AI" comment, I've left a few of my own and gotten thank yous in reply.
> You should add more AI to your life
I hope you can see how this is not a useful suggestion.
To continue your analogy, that would be like if Gmail got rid of spam filters, and then told people to stop complaining endlessly and manually copy every email they get into SpamGPT to ask if it's spam or not.
Actually, when 90%-95% of articles posted online are AI slop, it's even more useful to identify those which aren't.
When the signal/noise ratio is too low, having an indicator of signal is tremendously useful.
I think the criticism also signals to the submitter or idk co-author? That the article isn't valued
It seems better if HN users poke around and decide for themselves what the articles and threads are about, and which are interesting or not.
However, the second part of the problem - the AI comments, that's really what kills the discussion and eventually the community. If I have trust issues that I'm talking to a bot but not a real user, I am not going to engage further. I might or might not flag the account, but at some point, I'm going to be tired of reporting if everywhere around me it's just bots pretending to be humans. I think this is the more serious problem that needs focus.
In my experience, sloppy AI content almost always, sometime instantly even gets flagged out.
Some people already dismiss genuinely useful content solely based on the use of AI to assist in writing it - i am not sure what flagging would do other than to reinforce that prejudice.
I posted a couple of my articles here, and the one that got traction was generally well received (and also received some constructive feedback from those who acknowledged that is was AI assisted) - but it is evident across HN there is a vocal minority who outright dismiss content solely because it was "AI generated" completely disregarding the content itself. I appreciate this is personal taste, or LLM fatigue, or whatever, but its not really constructive.
If what you want to do is target the slop while not targeting the quality content, then that is what the voting mechanism already does. If people don't like something they can downvote it. Flagging content as AI generated is just a dogwhistle to those who want to downvote AI generated content. If anything, id rather see a rule that stops people commenting on stuff just to dismiss it as "LLM slop".
Its already trivial to avoid detection with fine-tuned humanisers [i] built on non-instruction-tuned models. That makes the flag mostly useless - or worse - a way of penalising any content you disagree with. I'd rather not hide what I am doing and have something that I feel reads well than hide it and sacrifice the message to satisfy a vocal minority.
[i] https://arxiv.org/abs/2605.19516
EDIT: downvotes, as expected. hope you see this anyway @dang. Downvotes kind of make my point for me.
Why is that a good thing? If you aren't going to bother to write something, other people shouldn't have to bother to read it, so please don't put it in the world at all.
Gatekeeping based on skill and motivation is good!
You are gatekeeping based on the medium, not the skill or motivation.
So why are you doing that by sloppyfying your own writing? If you want people to take interest in your writing, it should probably actually be your own writing.
Your prejudice/ableism in calling it "sloppyfying" is exactly the problem I mention with allowing it to be a flag.
Ultimately though this is the same debate as "should we allow genai code in codebases". High quality code lands naturally while slop is slop. Not much value in banning AI outright--the desire is predominantly to ban the slop.
Maybe the tag should be [slop] rather than [genai]...
I guess my response to that is we want the most interesting threads.
Single page info-graphics and Awesome-Lists come to mind.
Hell, one or more of PG’s books has one or more baysian generated texts presented as poems.
Congratulations on your recursive ascension, too
Also, votes and flags are personal, even intimate data. I can't imagine publishing it.
I know several people that have difficulty writing but are still f*cking smart. So what, they should refrain from using AI to help them because the AI police says so?
But yeah, let's continue classifying people based on their outer qualities and habits... History showed us were this leads us to.
And here an em dash -- to freak out the AI police.
Do what other people do when they cannot do something: don’t do it?
I can’t draw or do art and I don’t have an AI generate art for me so I can LARP as an artist.
You can just… not do things? Nobody is putting a gun to your head to make you write blog posts and post them on HN.
PS that's not an em-dash – this is.
The issue is that AI doesn't actually help them; it corrects grammar and removes voice. If you're trying to sound more professional/convincing/engaging, AI usage undermines that.
A strategy for people who feel they can't write is to just keep it short and sweet. Intelligence is such a potent force. It shines through.
What humans want most is to hear the voices of other humans.
Better is simply don't publish slop. If someone spends time refining and editing their idea with AI assistance and are happy with the quality of the finished product then I say go for it.
It's not like all human writers produce content that is worth reading.
For example, if I quote a GenAI response –even in criticism– (See what I did, there?), it can get flagged, and result in a shadowban (has happened to me -lesson learned).
But a good use for LLMs, is as a copyeditor. They do a great job. Some unedited stuff is so bad, I'd rather read slop, any day.
The problem is, what's the threshold? If they just fix a few typos and misspellings, that's fine, but what if they offer more substantial changes? How much text must change, before we can legit dismiss as "slop"?
Also, what if there's a significant GenAI component, but the article really is something that we want on the HN frontpage, because of its content?
I don't think this is true, at least not right now, and in a way I'm actually thankful for it.
The frantic rush to chase the only potentially profitable use case for LLMs found so far (writing code) and the resulting focus on coding RLHF means models are actively becoming worse at sounding like humans.
This is my favorite example, and it's already relatively outdated: https://progress.openai.com/?prompt=10
This is totally tangential to your point, but what is that page? It's not your usual link to an algolia search. Is this already part of some sort of manual tagging system? Clicking on the first one, these comments don't seem to be moderated. Are you using these complaints to help detect AI generated content? I think the existence of that page just leaves me confused on whether you actually want people to comment like this or not.
Keep in mind that we have a REPL over here and can make any link do anything!
Generally I'd appreciate a top level comment from the submitter saying this is LLM generated but I read it and found it interesting because x,y,z because I'd rather read slop someone vouches for than slop someone hasn't
Many are quite a bit more subtle, like this: https://news.ycombinator.com/item?id=48844062
The more subversive undercurrent is interesting to me. People intentionally fucking with someone's bot, burning tokens for the lulz.
(I already flag submissions I think were written by AI.)
just for a while :)
1st: the presumption that AI generated text is actually unsuccessful, rather than proliferating broadly unchallenged today
2nd: the disposition that negative attitudes towards AI text are unjustified discrirmination, rather than working as a strong latent predictor of low-effort content
3rd: the assumption that human writing is reliably doled benefits, rather than some poor proxy of it (winning the social contest for claiming authenticity)
Operationally, only a very small minority of humans actually successfully identify AI generated text at rates ≥ Pangram. People discriminate against the label of "AI", but mostly fail to vote accurately. It's not uncommon to see bots abusing this gap for their own success -- accusing humans, sympathizing with generated profiles... FUD environment where people routinely get away with dismissing true accusations.
For someone who is mediocre at detection, this would structurally feel like an unhinged, unjustified bias: look at all these good posts, these honest people, getting undermined by discrimination...
> 1st: the presumption that AI generated text is actually unsuccessful, rather than proliferating broadly unchallenged today
I can only talk about HN. If you think this is proliferating broadly on HN itself, I'd like to see such links. Assuming it's "broadly", they should be easy to find.
> 2nd: the disposition that negative attitudes towards AI text are unjustified discrirmination, rather than working as a strong latent predictor of low-effort content
I don't understand this bit.
> 3rd: the assumption that human writing is reliably doled benefits, rather than some poor proxy of it (winning the social contest for claiming authenticity)
Sorry, but I don't understand this either.
> only a very small minority of humans actually successfully identify AI generated text at rates ≥ Pangram.
That's not what we seem to be seeing. I do agree that there's a wide spectrum and a lot of wrong guesses.
> It's not uncommon to see bots abusing this gap for their own success
I'd like to see specific links of this on HN itself. Again, if it's not uncommon, those should be easy to find.
Edit: these are just gut-feeling numbers but we've had many dozens (perhaps hundreds) of email exchanges with users about this to date. Usually when we ask them if their post was genai, probably 90% say "oh yes, sorry, I didn't realize you had that rule". Of the other 10% or so, when we ask further, it usually turns out that they used some other tool (a grammar checker, a machine translation, some kind of toucher-upper, etc.) which left LLM imprints on their post. It's pretty rare for someone to insist that they wrote the text entirely by themselves with no LLM or similar tool. Of those, I'm sure some are false positives.
There have also been a few instances where I didn't fully believe them, mainly because they'd already posted other things that got classified as genai. Fortunately this is rare. The vast majority of users who we have this sort of conversation with are sincere, legit users who don't know much about HN. Our hope is therefore to convert them into informed community members!
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?
One of the main reasons that I (and I assume others) am here is because I can discover interesting content. It is true that there is a lot of spam in the internet but if I wanted that I would be in x, linkedin or sth. My problem with AI right now is that I consider machine generated content low quality one, and I would like to be able to decide if I want to read such an article without having to waste time before I realise it is ai generated.
> if an AI generated article is indistinguishable from a human written article and it is accurate and interesting, do you care who wrote it
Ideally I would like to know regardless. Practically in such a hypothetical future scenario it may be impossible, but I think that not doing sth right now because of some hypothetical future that may or may not happen is not a very good argument. Right now the AI content is pretty distinguishable, and if one took the time and effort to make it not seem like AI then at least that text contains some more human effort.
I care about the fact that I’m interacting with a human. If an LLM generates content that reads like a human and it isn’t disclosed it is deceptive by nature. It doesn’t matter to me that a human typed the words on their keyboard, they could use text to speech or whatever. But I care that a human crafted the content. A LLM doesn’t just type the words, it takes style, tone, voice decisions, in addition to the choice of framing. That’s everything that matters about human to human communication
having said that, i'd still much prefer a norm of including prompt history with the article, or the codebase for that matter, so people can choose for themselves :)
On the extreme end you have recipe blogspam, with the tropey life stories and Amazon affiliate spam that precede the meat (pardon the pun). If that pre-filler is inconsequential, why not have an LLM parse the semantic markup - which 99% of recipe sites use for Google SEO - to produce custom content for you? Like a history of the ingredients, or better instructions with an equipment list tailored to your kitchen and timings/temps for your specific oven?
I can see there being a middle ground where you publish some sort of context that a user’s LLM can generate the content from. Is it fundamentally any different from dramatized non-fiction? Think of books like Operation Mincemeat where the facts are thoroughly researched, but the author uses a lot of artistic license to tell the story.
But to your point, if there is back-and-forth, I wish people would pay more attention to the obvious traits of LLM writing. Not everything needs to read like it’s a snappy editorial, but I sense we’re still in the early days where people have been handed this technology and are simply excited that they can pump out 1000 words in a coherent narrative.
Regarding llm writing, it is disappointing. Especially as people could use it in the “opposite” direction, using the tirelessness of the llm to experiment with the communication with the aim to make it as clear as possible for the reader…
Also, the information density plummeted because of ads, not because it was demanded by readers.
Two showstopper problems with this:
1. Many of "us" don't like interacting with LLMs more than we have to, and view reading as a pleasure that would be extinguished instantly at a chat prompt.
2. It lazily implies that the essence of an article is so easy to communicate that it can be compressed into a prompt. There's a reason for titles, subtitles, open sentences, hero images, paragraph openers, etc etc: they pull you in, establish tone and voice and so forth. Where are the prompts that can entice a reader to even skim it much less plug into their favorite chatbot?
Yes, because I have an interest in traces and mechanisms of human connections being preserved.
Like, of course it matters who wrote it - an author imparts parts of themself onto their work. That’s what makes writing that’s worth reading.
I really hope there's a huge bot army pretending to be those people, but alas, you might be right and the reality is worse than that.
Well, yes, because of that "accurate" part. Humans, in general, actually attempt to validate what they publish. I don't feel the need to treat literally everything as a possible hallucination when I read an article published by a human. Less cognitive overhead makes for a more pleasant reading experience. In other words, I trust a human to be accurate most of the time, and I trust an AI to be accurate only some of the time, so there's much more to verify.
I simply instruct my RSS Reader to fetch articles only from blogs which I believe to be high quality.
The era of people caring about knowledge/learning seems to be dead. At least in the way we used to, because a lot of what we needed that for can now be done by the LLMs.
I do.
> We want to avoid low quality, not AI generation, right?
We want to avoid further dehumanization of the already semi-dehumanized humankind.
I don't think this is true. I think most of the people who post LLM content without editing and without even style guidelines are simply of the opinion that the resulting text is genuinely of an acceptable quality.
Agreed with the rest of your post. We should think of this in terms of quality of the result, not the process.
It’s the same reason recipe websites don’t honestly label that they ripped off other recipe websites: it would train users to go to the source.
For example, consider this project where I also complained about atrocious writing: https://news.ycombinator.com/item?id=48876506. I think this is really just someone who wanted to "create" something, I don't think this is in the same category as recipe/blog spam. I don't think the person who put this online thinks that this landing page is worse than any other programming language website.
I don't disagree with that assessment, with one extra detail: for many of them their bar of “an acceptable quality” is too low. No higher than “sod it, it'll do” and often lower. To too many, generated content is seen as a gateway to maybe accidentally becoming some sort of influencer, a way to get themselves out there with little or no care about whether their contribution is actually useful overall just as long as it garners a bit of attention.
I don't want to waste time hunting through the extra piles of “it'll do” or worse to find the remaining nuggets of actually good writing, be they entirely human made or created by humans using LLM assistance. There are at lease a few people of agree with me on the matter.
The problem with judging on quality is that you often have to at least start reading the slop to realise and move on to the next thing. Rating systems and flags may help somewhat, but they will all eventually be gamed so it will become yet another battle of attrition: people and systems trying to filter out the crap while other people and their systems finding it more profitable to spend time breaking the filtering mechanisms instead of producing content good enough to not actually be filtered by them.
As for the rating system, we already have many people who first check the comments for things like "AI, don't bother". I think that's OK; it could be better by quoting two or three sample sentences so that others could quickly judge how bad it is. I don't think it needs to be more formal than this.
Right now, in mid 2026, many of believe that LLM prose is not as good as human output.
The deception is one aspect. AI for spellcheck and proofreading is one thing; AI writing the content wholesale in the first person, which you pass off as your own, is another. I also don’t care for the Tailwind-styled demo sites that have flooded Show HN.
The sort of content that people react viscerally to (as slop) is rarely that subtle. Again, no problem with careful and subtle use (see CGI in movies), but when I see Claude-isms and snappy section headings that nobody used to use, I close the page.
You can/should write for you, to get experience writing, to present a portfolio, to share things you find interesting, a photo a day, whatever. Why does that have to die?
I can't believe no one had posted this yet.
Whether you agree that a con artist is only a criminal if he gets caught, or you think he’s a criminal the whole time, surely you can see why many people might want to know if they are dealing with a friend and not something simulating a friend.
if it were, 90% of startups would be illegal and they founders jailable
but we are not even talking about that, most authors do not claim LLMs were not used
That was a two-pronged thing. As well as the blog-spam drowning out what good blogs there were, a lot of people that used to output that way moved to centralised platforms that then either tried to lock their content in, or surrounded it with ad-tech bothers, or both.
> if an AI generated article is indistinguishable from a human written article…
Quite a bit of AI generated output is indistinguishable from crap human writing, often mediocre human writing, and that is actually part of the problem. A lot of people bang something out of their LLM/agents of choice and not bother taking the time needed to lift to good or excellent, so there is a bulk of mediocre or worse stuff flooding the medium because it is so easy to produce. But do we really need 100+ mediocre articles and a few good ones on a subject where there would previously have been a few mediocre ones and still a few good+ ones? It makes the genuinely worthwhile content much harder to stumble upon, especially as the ones taking the lazier approaches to writing seem to be making greater efforts in self-promotion with the time they have saved!
> if an AI generated article is […] is accurate and interesting, do you care who wrote it?
Firstly: maybe not, or at least less so. But I'm not convinced they can be trusted to be that accurate. I don't trust the average human that much either, they are often authoritatively wrong too (which doesn't help the LLMs, which have been at least partially trained on authoritatively wrong content), but I trust good human writers (who may be assisted by LLMs these days, like they've been assisted by spelling & grammar checkers for decades) above other human writers and content almost entirely drawn from LLMs.
Secondly: Yes. I care. I'm still smarting that after years of people being fined heavily, banned from things, and even imprisoned, for copying content, we seem fine with the big companies pirating whatever they want for training purposes⁰. Yes, some fines have been paid, but a few hours worth of income isn't even a slap on the wrist given how much funny money is being sloshed around the sector ATM, and do you know any content makers who got a share of those fines? I'm still trying to avoid being a part of that, to the point of being willing to commit career suicide¹, so yes, I do care who/what made most effort on creating the article. Given a choice between human, human with LLM assist, LLM with minimal human editing, and purer slop, I would morally prefer something from as close to the “just humans” end of the scale over anything else, even if the quality is practically identical.
--------
[0] I wonder how the corporates would take the “we only downloaded and used it ignoring the licence for model training purposes” point being used about their output. After all, I'm only using it irrespective of your licence to train the intelligence model that sits between my ears!
[1] The current corporate overlords have as much said “get with AI or be left behind”, and I'm sure being left behind will involve being PIPed out to pasture. I might be able to argue “the difference is enough to be considered a material change in roll, so you can't force that under UK employment law nor sack me for not playing ball”, but the same change-of-roll argument just opens up the possibility of redundancy instead², which is only marginally better. In either case I'm done for working in development with my current attitudes.
[2] this interpretation, if not legal pigswill entirely, also makes “we need less legacy developers and more agenic developers, your legacy roll is therefore redundant, do you want to change to an agent-based job or do you want to leave?” perfectly legal.
The idiom is someone says “tell us how you really feel” (which he’s rephrased here) to express (I think) that one’s expressed opinion is far too dramatic in its opposition of some thing. It’s meant to be read ironically.
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.
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.
If there is a great post on a topic and the author used AI when generating it, what’s so bad about that?
Is the goal of this site to remain high-quality, or to rely on random, anonymous users to discern quality? Consider the possibility that the two options might be incompatible.
FOMO-suffering hype junkies do care if there's enough fellow roof jumpers to join to and sing something like "well, define high quality first" in chorus on their way to the ground.
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.
Hacker News adopting such a feature would likely do more harm than good.
The issue is that label has expanded to anything that's weird/unusual, and the consequences of being accused of AI are far more severe than the consequences of a false accusation so people do it frequently. It is absolutely a vector for harrassment/trolling.
One subreddit I follow now bans people if they make an accusation of AI-generation that's weak/disproven, which is a rule I like.
https://github.com/alvi-se/ai-ublock-blacklist
https://github.com/laylavish/uBlockOrigin-HUGE-AI-Blocklist
https://github.com/Stevoisiak/Stevos-AI-Blocklist/
That's a good example of the exact problem with such a broad stroke rule.
Depends on your use-case, but most people don't need content used in network nuisances like YC AstroTurf posts. =3
Even scoring posts does provably change peoples behavior, and studies showed it tended to make people more punitive in their conduct with strangers.
At bare minimum, YC should have bot CAPTCHA protection on posts. As unleashing chat bots on users is often disrespectful, and just cows people out of participating in conversations in good faith. =3
Are you implying the front page is botted/manipulated?
Throwing up a bot challenge would probably improve signal-to-noise ratios, but who knows for sure. =3
Quoting myself: https://news.ycombinator.com/item?id=48063759
>> Every post that reaches the top of HN will have at least a few comments saying "This is LLM!"
It has become a proxy for "I don't like this article, so it must be a LLM"
To me, it feels like lazy karma farming, as these comments often do get a few upvotes.
And of course, accuse a 100 posts if being LLM, you are guaranteed to be right at least once, then like astrologers you can claim success.
Is there anything we can do to discourage this type of lazy and low effort posting?
Conspiracy minded responses are low value, and yours is especially so considering HN bans AI comments.
Tagging an opinion different from yours with a 1967 CIA axiom is of even lower value.
WTF? What do you mean by "hosted"?
I know it's the internet's happy place to blithely accuse others of monstrosity and then derive from that how nobody other then themselves has any "morals or ideals", but this one sticks in my craw.
Here is me referencing it about 11 months ago and I have older comments that reference it. https://news.ycombinator.com/item?id=44864419#44866331
I guess HN likes to scrub things that make them look bad instead of confronting the reality that they helped launder and "humanize" genocide.
We don't ever delete anything from HN (except in rare cases where a user asks us to do so for privacy reasons). If we had deleted something as controversial/inflammatory as what you're claiming, it would easily be exposed; the HN API is public and many people ingest and store everything that is posted, so someone would have a record of it and could easily dig it up. That's aside from any cached copies in The Internet Archive, any screenshots of it, any tweets about it, etc. We never even consider trying to hide anything that's happened on HN, because it would so easily be exposed if we did. If something bad happens on HN, we own it and work to improve.
Edit:
The thread you must be referring to is this, in January 2024:
ICJ orders Israel to prevent genocide in Gaza, stops short of ordering ceasefire - https://news.ycombinator.com/item?id=39143043 - Jan 2024 (1401 comments)
This comment, which appears near the bottom of the thread, is from someone identifying as an IDF soldier and sharing their perspective: https://news.ycombinator.com/item?id=39147721. That user posted 9 comments in a thread with 1401 comments. Most of the replies to them were hostile but attracted no moderator action. Some of your replies attracted user flags and comments from dang due to breaking the guidelines. This is routine moderation. The guidelines apply, regardless of the topic; otherwise there would be no point having guidelines, and we'd never be able to have discussions about any difficult topics at all.
The main topic and the overwhelming majority of comments in the thread were critical of the IDF.
For the record, this is the comment that dang replied to: https://news.ycombinator.com/item?id=39158222.
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.
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.
They are driving (lol) towards the sale of a computer with a model on chip without a programmer salary and benefits adding to cost and resource use (all the tools of the trade need to be stored and copied around).
This is capitalism, not honorific obligationism! Saving your job is no one else's burden! Disrupt! (YC crowd 15 years ago). Oh how the turn tables.
Move the context of the code to the presentation layer and eliminate state by keeping the hardware focused on computing geometric transforms. Biggidy bam label the data on the display and save both geometry and labels
There's no reason for biz logic context to exist in code.
Just stokes addiction to a false sense of empowerment and social contribution
Downvote away to hide from ideas and opinions you don't like.
Conservative people are the only people so dysregulated by ideas they don't understand or that do not validate prior experience and embedded biases they just try to hide them from view. Not the kind of people whose judgment I concern myself with in any meaningful way
I agree. We've had large and thriving online forums for at least a decade before upvote systems became ubiquitous and things were just fine.
yeah keep chatting with the sycophantic psychosis-inducing LLM chatbot, I think you'll have a better time.
LLMs are mass theft of intellectual property. I didn't need them to "drive" this computer for the last ten years.
they are different systems created for completely different reasons. the idea of calling them all "property" came from lobbyists who wanted to make people think they are natural rights that should be expanded and protected at all costs, instead of tools made for a specific purpose.
thats how we get todays system where publishing anything even without a copyright notice gives you life plus 70 years of total exclusivity, but also if you signed a bad contract with your employer they will act like all your work belongs to them even if you did it on your own time and threaten you with lawsuits if you try and fight. its bad at protecting artists (just look at the music industry) and even worse at protecting the public.
what llm vendors (not the models themselves) are doing is not really theft because you cant steal something if the original owner has it the whole time. the bad thing they actually do is train llms on everyones work and then claim copyright over the result. that is the real injustice. and its why i think open source === ethical when it comes to ai. if you take you have to give back.
the only true permanent solution is a full rebalance. all "ip" rights should be for commercial use only and last 25 years or less. yes that means piracy is legal as long as you dont make money off it. unpaid royalties are a problem for real people but pirates only hurt corporations. i know what side im on.
"intellectual property does not exist" -> goes on to define various legal instruments for protecting intellectual property.
"you can't steal something if the original owner has it the whole time" Is this fifth grade? So if I go over to your house and trash it, it doesn't matter the house was under your possession the whole time.
No the rebalance is to sue these companies into the dirt.
I'm glad the idea is picking up steam.
IMO, the post title should get "[AI Generated]" at the end if enough people flag the content as AI.
Personally I recently published a white paper on an idea for democracy that has been bouncing around in my head for decades. AI helped flesh out the idea and write the pages of text and structure the whitepaper.
I'm not an academic so I probably would never have fleshed out this idea without AI but I think it's better to have the idea published so it can be seen than to have remained bouncing around in my head until I passed away.
I'd agree that an academic version of my idea written by hand would be better. But a mostly AI written version is better than the knowledge never being published and I still spent a good two days on it making sure I agreed with every paragraph and fixed every issue I could find.
Considering this in general I think the amount of human effort that goes into creating something is probably the right measure of it's quality over if AI was used to assist in it's creation.
- LLM for helping with structure: sure. feels to me like more of a motivation problem to commit a first version to paper so you can start editing and moving bits around, and the LLM's role is being a conversational partner who cares and talks to you, but whatever works
- LLM for writing the pages of text: I would almost certainly prefer the version where you just write the key aspects (your insights) and not have to wade through pages of prose to find them
- LLM for fleshing out: hell no. if you want to lecture and bore people to death, this is the way to go
Turning insights into elaborate documents is relevant once this makes it to legal codification (compare "tell people how you use their data" with the relevant articles in GDPR)
But I'm not sure there's a great way to handle it. Flagging works as AI generated is good in theory, but it's become a bit of a witch hunt online, with plenty of human created pieces getting wrongly flagged as AI generated due to using em dashes in text, having the hands drawn awkwardly in artwork, or using other stylistic traits that AI content overuses. I fear half the site could end up flagged as AI-generated, just because a lot of people are hyper-vigilant about such content and assume the worst for everything.
At the same time, there's not really much of a way to incentivise writers to flag their own articles here, since revealing that a work is created by AI is a great way to both kill your credibility and drive away about half the people who'd otherwise enjoy it. So, if it's up to the authors, the incentives are for them to lie through their teeth.
I also don't really trust many AI detection tools, since they usually use AI themselves and misflag a lot of content.
But I don't think the site should change that much. Maybe add an option for AI content in the flagging system, and assume good faith for submissions in general. I'd rather not see the site get too paranoid or restrictive over this stuff.
https://soitis.dev/comments-owl-for-hacker-news
i just manually hide anything ai related.
first pass: hide anything related to AI on front page.
second pass: is there anything interesting left on the front page?
sadly the answer to the second question is often no nowadays :/
If you can easily detect the pattern of LLM writing here's something for you to look at: I was reading some pre-consumer-LLM papers by AI researchers and founders and the way these are written are incredibly similar to existing LLM prose!
Works on a personal level, unsure if this would work well in practice. Maybe just a tag is enough, so people can conclude for themselves.
I know there are extensions out there that are doing the democratic part right. Mostly YouTube-related extensions like DeArrow and SponsorBlock
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)
It's not discrimination to ban a practice which has been proven to be harmful. AI slop is harmful to readers, as well as being harmful to produce. Whatever issues one might raise, for instance someone who is not a native English speaker, we can find alternative solutions which are better for the environment and better for human minds.
I have a hard time finding these communities
AI isn't a higher power than HI and it never shall.
It is the author's and writer's duty and sole responsibility to tell viewers/readers that their respective works are AI-generated or how much percentage of it.
I don't care if a human, an AI or a cat wrote it.
We can do it already if we ask the AI companies to use one of the special whitespace characters instead of ascii 0x20. It would also help them avoid the problem of feeding their training loop on generated data.
The day you get your "AI generated" flag, I'll want my "lousy with mistakes", too.
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
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.
> 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.
And im sure this was designed in order to encourage positive discussion
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.
My entire position: I'm not interested in reading text that sounds like it was written by AI.
I think this is a fair position - if the author did not even take the time to read their own article, why should they expect others to read it?
If they did read it, they would have spotted instantly how AI it sounds.
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.
I do think the dead giveaways (em dashes, it's not X it's Y, etc.) are annoying to come across repeatedly. A person not bothering to remove these tells feels 'low effort' to me.
I used to love reading HN for the handwritten articles and handmade projects, but in the LLM era the quality has deteriorated significantly. I find myself flocking to other message boards where LLM content is flagged, discouraged, or banned.
However, a lot of the academic or technical posts on this site have turned out to include AI hallucinations (a problem that is not just on HN, in fairness). If an article or paper contains nonexistent or obviously, blatantly wrong citations, then I feel very strongly that such content is disrespectful to the HN audience, certainly to me, because I've just wasted my time trying to take the author seriously only to discover that the argument was founded on a hallucination. In a sense, it doesn't matter to me whether the hallucination was due to AI or not -- if a person puts in bogus citations, AI-generated or otherwise, my view is that their article or paper fully deserves to be flagged so that we don't waste other people's time.
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 in their making.
Wait I'm not sure I understand the cause and consequence here if you're saying you use an LLM but your articles aren't slop (note that those are rough synonyms). Do you mention in your submissions you subscribed to or used Claude? The mere existence of the subscription and using it for unrelated things won't be why people flag, I can only assume they notice the quality level and/or style
Not a single word of my articles was written by LLMs. I used them for proofreading and critique.
Writing articles by hand is useful for me since it allows me to crystallize my knowledge. Writing it down for my own future reference essentially prevents me from forgetting in the first place. LLMs accelerated me so much I'm actually trying to slow down so I can let my recent achievements sink in a bit. Writing is how I do that.
> Do you mention in your submissions you subscribed to or used Claude?
I don't use Claude for my articles. I use Claude for code review and design. Implementation is almost a footnote, and even that gets significantly rewritten during my numerous human review passes. I still write code by hand too.
That's for the projects I actually care about. I do have some explicitly vibecoded projects, and I never made a secret out of it. Projects that wouldn't exist at all were it not for LLMs making it easy to execute. I personally review those too, just not as often.
I don't add AI models as co-authors to commits of my own repositories. I'm not in the business of providing free advertising to trillion dollar corporations. If other projects require it, I do so as per their rules.
> The mere existence of the subscription and using it for unrelated things won't be why people flag
It will. They will zero in on things like CLAUDE.md and dismiss your entire project as slop the second they see it.
I know because I asked people directly. Response? If an LLM touched your project, then your project is slop. Verbatim. And that's the polite reply.
Trying to establish some simple criteria for sloppy vs good work nearly got me kicked out. People came at me with literal "I know it when I see it" nonsense. Even Claude does a better job of defining "slop" than that. I find it hilarious how these rules are based on vibes. People just get LLM vibes out of other people's projects, and then dismiss them.
> I can only assume they notice the quality level and/or style
I'm not sure they even saw my project. It's just pure prejudice. AI detected? Slop, no need to even read it. If they actually read my work and thought it sucked, fine. In fact I'd be very interested in knowing why so I can improve. I refuse to accept these predetermined judgements though.
Getting outright called "LLM pusher", "slop fetishist" and "clanker lover" is pretty fun too. It's not the overt racism-tier insults that bother me, it's the fact not a single person seems to have gotten banned over this. One could be forgiven for thinking they opened a 4chan tab.
LLMs essentially cured me of this particular problem, to the point I got addicted to building things and had to actively dial it back a bit because it.
As such, I consider LLMs to be assistive technology, like screen readers. LLM opponents are displaying some serious ableism.
AI slop is AI slop.
A good article is a good article. Doesn't matter who wrote it.
Obfuscation of facts is the major issue with AI articles that should be bopped.
The problem with these texts to me is that the parts that are information-dense are often not real and the majority is not information-dense. It’s just filler text of a sort that’s pointless “35% ram. 3x throughout. No latency trade off. That’s the whole point”. Okay, what’s this random “that’s the whole point” added there. Useless.
I know it’s passé to say “HN is becoming like X” but this is pure LinkedIn slop. Someone publishes pure bullshit and their fan club posts a bunch of likes and “I’m so excited to see this. Great post”.
Voting systems can be gamed and as HN becomes bigger and bigger it'll start to attract unsavory audiences who have an agenda.
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.
Is it purely just a "human supremacist" desire that fuels the motivation to ban or block such articles?
It's not about the end output. End output is already good IMO. Now, the sentiment is "if you don't give a damn about writing what's on your mind, why should we give a damn about reading it?"
It doesn't matter how bad your writing is. You have a flavour to it. You write and make it better.
Not to mention, everytime I write something, my thoughts get clearer and clearer. Vibing an article kills this process. It's about the unknown unknowns. You don't even know what you don't know until you start writing. LLM doesn't help you here. It might help you get some likes.
2. judge content not by its cover and think.
If we are supposed to not judge the content by its cover, then don't put a cover on it. It's that simple. AI adds no value to the content, just some noise that makes it harder to reach the actual information/emotion.
Yes, and those differences have an intrinsic value. As an example, my comment's "cover" communicates my discontent and frustration with "AI tone". As a matter of fact, my "cover" here became the discussion itself. Thus, my comment's "cover" is not actually a cover but a part of the communication. In contrast, the "AI tone" is just a cover to make a writing look superficially more professional without adding any value to the communicated idea or emotion.
2. Labeling the ones that are honest about being AI generated punishes honesty and rewards lying by boosting AI generated articles that say they are not AI generated.
Would you skip articles because it's written with a text processor? You need to have it written bit by bit by fusing it directly in memory?
AI / LLM is the new word editor. Get over it.
What I find really annoying is all the comments that pretend to see / detect AI slop... with lot of false positives.
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.