Should we summarize HN submissions?

No. HN is a site whose primary purpose is to link to other sites (and maybe promote VC investments). If you don't want to click links to other sites, don't come here. If you want summaries, go to Reddit. HN is different on purpose.

> If you don't want to click links to other sites, don't come here.

I'm afraid that you've completely missed my point. I do want to click links to other sites, which is precisely why I come here.

I said, for example, HN readers "avoid clicking on articles that they would be interested in reading" due to article titles that are uninformative.

Nobody has the time or desire to click on every submitted HN article. (I think HN may average a submission every few minutes.) The point of summaries is to direct readers toward the articles that they would be interested in. This would also tend to increase the value of the submission comments, because currently HN readers click on articles that they regret clicking on and come back to the comments to complain about them.

A summary is not a replacement for reading an article. A summary tells you why you should read an article. Call it a "pitch" rather than a "summary" if you like. The problem is that article titles themselves do not make good pitches, but that's the only current pitch we have.

No. I'm not at all interested in an editorialized summary of an article.

Either the title entices me to read the article, or it doesn't. Both are fine. Besides, I'm here for the discussion, not the articles.

If you want to have summaries, create a browser plugin that automatically summarizes the articles. Do a write-up about the plugin and submit that as an article itself. I'd read that! You then scratched your itch and others might benefit. Without changing HN itself. Win-win!

> Besides, I'm here for the discussion, not the articles.

Why even bother having URL submissions then?

> create a browser plugin that automatically summarizes the articles

I assume that by "automatically" you mean by LLM?

No thanks. The only reason HN interests me is that it's curated by humans. Otherwise, I could just skip HN entirely and have LLMs pick out lists of articles for me to read. In fact I could skip HN users comments and generate those by LLM too. Who needs humans, right?

We've always discouraged summaries, TL;DRs, and now, LLM-generated digests of articles, and we don't seen any reason to change this position.

The main reason is the same reason for insisting that the original title be used (unless it is baity or misleading); it's not fair to the author or the community for someone to take it upon themselves to re-phrase the title or the article’s content into something different from what the author wrote themselves.

The community won't know – without reading the full version – whether the rephrased version is an accurate reflection of the author's original words or ideas. But some commenters will write comments in response to the rephrased version, rather than the original, which causes the discussion thread to be poisoned.

We prefer to maintain an expectation that people will read the full article before commenting. Of course, many won't, but at least if any misplaced comments are based on a mis-reading or non-reading of the article, it won't be due to an inaccurate summary that is tacitly endorsed by HN.

> it's not fair to the author

To be honest, as an author whose articles have been submitted to HN a number of times, I have a hard time believing that HN actually cares about being "fair" to the article author. To the contrary, HN has become notorious for unfairly and ignorantly tearing down article authors. You must be aware of this, yes?

I think it would be refreshing to have a summary from the submitter who, more likely than other commenters, read the article and appreciates it. As an article author, I say that I'm least worried about the submitter being unfair to me.

> We prefer to maintain an expectation that people will read the full article before commenting.

The community missed here is the people who not only don't read the article but don't comment either, who completely ignore the article, because the article title was uninformative.

I gave an example in another thread of a submission whose title is so vague as to be practically meaningless to those deciding whether to read it: https://news.ycombinator.com/item?id=44945966

In the absence of a useful title, the only way for someone to determine whether an article is worth reading is to check the upvotes and/or the comments. But that, I think, makes the anti-summary attitude ironic, because upvotes and comments don't come from the original author. Indeed, one of the reasons that many people go to the comments before reading the article—and HN users explicitly say this—is that they want to determine whether the article is worth reading. From my perspective, a brief summary or "pitch" if you will, would encourage more people to read the article and/or discourage people from reading an article that they come back and complain about having read.

And I would say, by the way, that one of the main reasons that submitters editorialize the article title is the title itself is not particularly informative.

> tacitly endorsed by HN

I'm not sure what this means exactly. In one sense, every submission and comment on HN, even the comments from people who haven't read the article, or comments that are otherwise inaccurate, is tacitly endorsed by HN, except perhaps those that have been specifically moderated by you. Indeed, you might even say that HN tacitly endorses every article that is linked by HN!

In another sense, though, a submission, and the proposed summary of the submission is not endorsed by HN but rather by the HN user who submitted it, whose username is attached to the submission. And I don't see a problem with that. This is always the case for a website with user-submitted content. Everyone can tell the difference between the site and the user.

By the way, while a moderator is here, the person who apparently summoned you here accused me of sealioning: https://news.ycombinator.com/item?id=44948008

https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...

Was that link supposed to make a point? I think it needed a summary from you. ;-)

As far as I can tell, HN receives a new submission on average every few minutes. Nobody can possibly click all the links and read all the articles. So, HN readers need some way(s) of determining which articles are likely interesting and which are likely not.

Imagine if HN submissions included only the URL and not the article title. That would be silly, right? But why would it be silly? Because then you'd have little idea what the article is about, unless the URL itself encoded words that explain it.

If we admit that HN readers need some kind of guide to the vast number of articles submitted, then I don't see why it's so far-fetched and seemingly unholy for the submission to add a brief summary.

I actually want to read interesting articles. I'm perfectly willing to "work a little", as it were. There are interesting articles that I miss, that I would read, if not for the uninformative article titles, and that's unfortunate, I think.

[flagged]

> I mean, you kind of asked for a response from mods by asking me to speak to the guidelines

I was requesting that you literally quote the relevant passage(s) from the guidelines or from other sources, which is simply copy and paste, not speaking for anyone.

You claimed that something was in the guidelines, but I read the guidelines and didn't find the thing you claimed, so direct quotes would seem to be the natural next step, not emailing the moderators, which feels like overkill.

> I don't think it's fair to call what I originally said an accusation per se, because I don't think it's necessarily up for debate what I was reminded of, as I am an authority on my own state of mind.

You can be reminded of anything you like, in your own mind. When you publish it, as a reply to me, it becomes for all practical purposes an accusation. I have thoughts about various people, including you, that I refrain from publishing.

> You yourself said you weren't finding it very entertaining, which to me seemed sarcastic and flippant

No, I was quite honest and sincere, about that and about the purpose of HN itself. The HN guidelines say that on-topic is "Anything that good hackers would find interesting", "anything that gratifies one's intellectual curiosity." That's more or less entertainment. Highbrow, perhaps, but still entertainment.

> If users and mods acknowledge your point as being worth discussing, yet disagree with your desired course of action, where would you like this conversation to go from here?

I understand disagreement, which is fine. I also understand that summaries are subjective (which I also think is fine). What I did not understand was your comments about "bad actors", "charlatans", and "sophists". This is a threat model that does not seem very applicable or realistic in the context.

>To be honest, as an author whose articles have been submitted to HN a number of times, I have a hard time believing that HN actually cares about being "fair" to the article author. To the contrary, HN has become notorious for unfairly and ignorantly tearing down article authors.

This reads as if you're confusing HN the site (and its goals) with HN the community (and the behaviors of those people). The ideal is that the site goals and the community behavior would be identical. We know that's not how things unfold in practice.

I'm not confusing them. But I would say, for example, that the HN guidelines and moderation, which come from the top, are focused more on encouraging commenters to respect each other, much less so on encouraging commenters to respect the article authors, who I guess are assumed to be not reading the comments.

I agree that the moderation is more focused on the commenters' behavior toward each other, but I disagree about the guidelines being focused there. Many of the guidelines are written so they are equally applicable to HN discussions as they are to authors (and their works). For example,

>Edit out swipes.

>Thoughtful criticism is fine, but please don't be rigidly or generically negative.

>Please don't fulminate. Please don't sneer...

>Eschew flamebait. Avoid generic tangents. Omit internet tropes.

Even so, I haven't seen any basis for thinking article summaries would change those community behaviors.

> The submission form says, "If there is a url, text is optional," and indeed most submissions with a url provide no text.

I think that's because most folks don't know how to do it or aren't able to do so. I don't know how some folks make text posts underneath their URL submissions, unless they are asking mods for help or using some kind of special post type like Ask HN/Show HN. Even then I'm not sure if that's all that is needed for users alone to make these kinds of text+URL submissions without mod intervention.

So if you are saying that this is a thing that is already supported but not used, can you link to such a post that you have made? As it is, I'm not sure that HN even lets you make posts like that.

Otherwise, I guess you can be the change you wish to see, but the guidelines argue directly against the point you're making, basically saying that commentary (which would include a blurb or summary) should go in a comment, not in the submission itself. No one is stopping you from summarizing articles, but just know that AI/LLM and other kinds of generated commentary is against HN guidelines to my reading also, so you would have to do that yourself. It seems likely that what you're asking for would result in more rehashing and less original content, but if that rehashing serves some kind of useful editorial function to raise awareness of content that is otherwise relevant and topical to HN and to its users, then your idea might have some potential merit, but I think it would be gamed by bad actors, and I think the bad actors would benefit more from this change than good actors and existing/future HN users.

I'm cautiously pessimistic on this idea, but willing to hear more about why you think it would help rather than allow charlatans and sophists to invisibly rewrite history and the present to bake in their own ideology and beliefs.

> So if you are saying that this is a thing that is already supported but not used, can you link to such a post that you have made? As it is, I'm not sure that HN even lets you make posts like that.

I think that HN simply makes your text the first comment. For example, here I linked to a social media post, and it made sense to quote the entire post to save a click: https://news.ycombinator.com/item?id=41556661

> the guidelines argue directly against the point you're making, basically saying that commentary (which would include a blurb or summary) should go in a comment, not in the submission itself.

Where do the guidelines say this? I'm not seeing it.

> AI/LLM and other kinds of generated commentary is against HN guidelines to my reading also, so you would have to do that yourself.

Definitely. I'm not calling for AI summaries. I don't want AI summaries.

> It seems likely that what you're asking for would result in more rehashing and less original content

Why?

> I think it would be gamed by bad actors

What do you mean by a bad actor?

> allow charlatans and sophists to invisibly rewrite history and the present

I honestly have no idea what this means in the context under discussion, especially the "invisibly" part but also the rest of your claim.

The submitter writes a brief summary of the submitted article. Everyone knows it's the submitter's summary. The worst that can happen is that the submitter writes a bad summary, you click the link, and the article turns out to be not what you expected. This already happens with article titles. Then you can come back and complain about the summary in the comments to the submission. I'm not seeing what the big deal is here.

Here’s an example of the kind of post I mean, with a URL and also text:

https://news.ycombinator.com/item?id=44935658

Your other points are answered in the guidelines, news guidelines, and/or comments by mods, which are also considered part of the guidelines.

> Your other points are answered in the guidelines, news guidelines, and/or comments by mods, which are also considered part of the guidelines.

If you'd kindly point exactly where, I'd appreciate it. Otherwise, I still have no idea what you're talking about. I have of course read https://news.ycombinator.com/newsguidelines.html

I have emailed the mods to help answer your questions, because I don’t speak for them or HN.

> I don’t speak for them or HN.

But you do speak for yourself, and indeed the mods don't speak for you, which I think leaves your own opinions unexplained and confusing:

> It seems likely that what you're asking for would result in more rehashing and less original content

> I think it would be gamed by bad actors

> willing to hear more about why you think it would help rather than allow charlatans and sophists to invisibly rewrite history and the present to bake in their own ideology and beliefs.

Summaries are subjective, so the subjectivity allows for glosses and elisions of the objective nature of the original content.

> Summaries are subjective

Yes, so what? The comments on a submission are subjective too.

I still don't know who you think the "bad actors", "charlatans", and "sophists" are, how they can "invisibly rewrite history", or what you think "gamed" means, aside from getting people to click the submitted link.

The entire point of my proposal is simply that a summary of the article, longer than the article title but still relatively brief, can frequently be a better guide than the article title alone as to whether the article is worth reading. That's it. So the histrionics about rewriting history and such just seem bizarre and inexplicable to me.

> the objective nature of the original content

Objective in what way? The submitted article may itself be an opinion piece, and often is.

Anyway, nobody would be stopping you from reading the article. Quite the opposite: the point of the summary is to explain why you should read the article yourself!

Here's an example: https://news.ycombinator.com/item?id=44945966 The article title is "What could have been". The article title is so extremely vague as to be useless. I have no clue what it's supposed to be about.

By presenting the summary at all, it is presumed to be accurate by good faith readers, but bad faith summarizers will take advantage of this knowledge gap to invisibly rewrite history, that is, they will present the summary as accurate and not tell you that they’re selective with the truth. The fact of the matter is the original content says what it says. Any summary is inherently subjective in comparison to the objective reality of what OP wrote when they wrote it. In this way, subjective summaries of the objective reality of the original subjective statements by OP have extra subjectivity due to the summarization. It has nothing to do with whether or not OP was making objective statements themselves, though that is also a factor in how much objectivity the summary can work with, but that is kind of a side point and not related to my original comment.

You're practically making this into a life or death issue, when in fact the issue is relatively trivial: should I bother to read the submitted article or not?

When a HN user submits an article link, the intention is that people will click the link and read the article. A "bad faith summarizer" is just shooting themselves in the foot, because if the summary inspires people to read the article, then they'll immediately discover that the summary was bad. So what's the point of the bad faith summary? Your hypothetical scenario doesn't even make sense. The URL is still there. The article title is still there. Everything that's available now would still be available.

Moreover, there's nothing now stopping any commenter on a submission from giving a bad faith summary of the article.

It's just like the TV guide. You've got a long list of channels, and you've got a bunch of show names, but you don't necessarily know what the show is about, whether it's worth watching, so you read the summary. Are you going to complain about the existence of the TV guide, because it's "subjective"? Hacker News is more or less entertainment, although this conversation is not very entertaining and getting less so as it goes on.

>The entire point of my proposal is simply that a summary of the article, longer than the article title but still relatively brief, can frequently be a better guide than the article title alone as to whether the article is worth reading. That's it.

Here's the line in the guidelines that seems to be eluding you in this whole discussion:

>It's implicit in submitting something that you think it's important.

The fact that an article has been submitted is the "better guide" the site uses to indicate "whether the article is worth reading." Tomhow's comment says why HN has not seen this to be something that should change.

It's implicit that the submitter of an article thinks it's worth reading. It's not implicit that other HN users think it's worth reading. Again, HN receives a new submission every few minutes. Not every article is worth our time; there's not enough time even if they were.

Thus, the point of the summary is for the submitter to explain to other users why the article is worth reading. Moreover, HN submissions are on a very broad range of topics, so even if an article is worth reading on a specific topic, readers are not interested in every topic. Some article titles are so vague that you can't even discern the topic!

>It's implicit that the submitter of an article thinks it's worth reading. It's not implicit that other HN users think it's worth reading.

This is what tptacek was noting in the search results he posted: Dang has made clear that HN is okay with having its readers "work a little."