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.
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!
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?
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.