Should we summarize HN submissions?
Submission titles are necessarily brief, not only on the author's side but on the HN side, whose title field has a limit of 80 characters. Some article titles are clickbait, and others are too clever for their own good, but even if the author is trying hard to accurately portray the content of the article, there's only so much you can say in a title.
The result of this situation is that HN readers end up clicking on articles that they would not be interested in reading if they knew what the articles were about, and they avoid clicking on articles that they would be interested in reading. As a workaround, some HN readers just go straight to the comments to see whether the article is worth reading. However, the comments tend to be hit or miss, ironically for that very same reason: a lot of readers are going straight to the comments without reading the article, so there's a kind of tragedy of the commons where a large % of the commenters are ignorant about the article. Moreover, even for commenters who have read the article, casually dismissive comments can rise to the top, upvoted by readers who haven't read the article.
What I'm suggesting, then, is that as a matter of standard practice, it could be very useful for article submitters to summarize the article in the submission text. Presumably the person who submits and article has read the article and is not dismissive of it, considers the article worth submitting. And if the submitter has not read the article before submitting, the absence of a summary would be at least a semi-reliable indicator of that fact.
I think it would be interesting to have expanded HN submission lists that include summaries inline, perhaps with the summaries limited to a certain number of characters.
4 comments
[ 2.8 ms ] story [ 21.3 ms ] threadEither 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!
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.
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.