Comment by vepea2Ch

5 years ago

That's really insightful, thanks for taking the time of sharing it.

That being said, without considering our own opinions on a given topic, you can easily reproduce the experience of visiting random HN threads and finding a well written dismissive post on top of the thread, no matter what the subject is, and almost systematically (and thus, when someone is interested in the topic, that's the first thing they see). This is hardly explained by the "monster neighbors shock" effect. If you agree with this observation, how would you explain it?

7 comments

vepea2Ch

Reply
dang  5 years ago

I'll offer two explanations. First, there's a contrarian dynamic on the internet: people rush in to express whatever they disagree with or dislike about something (https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...) and such comments tend to get upvoted because, for whatever reason, critical/indignant comments get upvoted. This is a weakness of the voting system (https://hn.algolia.com/?dateRange=all&page=0&prefix=false&qu...).

Second, though, I think you might be describing your own shock experience here. Not every thread starts with a shallow dismissal—some do, but actually most don't. (Moderation is a factor, because we downweight petty and indignant comments whenever we see them at the top of a thread.) My bet is that you're seeing these sometimes, and because they're shocking and unpleasant, they somehow expand into your experience of HN overall. That's a shock experience, because the things that strike us unpleasantly end up dominating our sense of the whole. I've written about this a lot, but in slightly different terms: https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que.... Trying to figure these phenomena out is an ongoing process.

  • vepea2Ch  5 years ago

    I see. It makes sense, thanks. I guess that's actually the problem that silo'd discussions solve (if you see only what you already like, you won't rush to disagree, and your comments won't shock people). Although, that's quite a depressing conclusion :)

    Out of curiosity, do you have any leads on what may replace a voting system for emerging insightful content? It sounds like a job for AI, but I guess any bias in it would be hated even more passionately.

    • dang  5 years ago

      I don't think AI has good enough taste yet. The best possibility I know of is giving users a higher-signal (than upvoting) way to single out a post that they think is particularly good. This would be like flagging, but for good things rather than bad. Actually, that was our intention when we created the 'favorites' feature and made favorites public—but it didn't work out that way, so maybe this is harder than it sounds.

      There's a bit more discussion about this at https://news.ycombinator.com/item?id=23157675.

      1 reply →