4 ms·
That's funny. I tend to skip the article and go into the comments. Comments tell me "what's wrong with the article", and in doing so, give you a good picture o
by richforrester 9y ago
That's funny. I tend to skip the article and go into the comments.
Comments tell me "what's wrong with the article", and in doing so, give you a good picture of what the article is about.
As for the "what's wrong" part itself: generally speaking, you can really tell by how the comment is written whether it's legit or not. That way, you kinda get two sides of the coin, and you can decide for yourself what's going on.
- aaron-lebo 9y agoI really just don't believe you can determine the truth content of an comment by how it is written, but maybe that's years of reading good sounding but factually inaccurate comments on HN and Reddit. There's an art to writing psuedo intellectual bullshit, and most people aren't capable of (or aren't as good as they believe they are) at judging complex issues outside of their domain. Would just as soon not give armchair phd a platform. There's some research (don't feel like pulling it up - bullshit?) that talks about how the decline of valuing expertise has had some negative social consequences, but that's probably obvious.
- sametmax 9y agoI pretty much do the same with HN.
- StavrosK 9y agoI think you two are thinking of different things, and missing an important point: Comments are only as good as the community that's writing them. YouTube comments are the kind of comments you don't want to read, HN comments are the kind you skip the article for. Whether comments are valuable very much depends on the kind of people you have writing them, and I believe it's worth trying to maintain a good commenting community.
- richforrester 9y agoI kind of agree and disagree. Yes we might be thinking of two different things. But I think it's not about the URL, but about the content. There's plenty YouTube videos that are intelligent, with intelligent responses to it. But that was kind of my point; by looking at the way a response to a piece of content is written, you can judge its merit somewhat accurately. For example, on a video about the way the Suicide Squad movie was edited, you might see: - "BRILLIANT 100% agree!" - "While I think you're right, there might be a couple of things you missed out on [...]" - "FU M8 DA MOVIE WAS AWESUM!!!1@2" Each of these comments tells you something different. And yes, some say more about the audience than the content ;)
- KGIII 9y agoSimilar, I skim the article and read the comments. In some cases, I just read the comments. Though, I don't do this on any direct media sites, like online newspapers. At Slashdot, chances are that I haven't read the article. I go straight to the comments and read those. Curiously, I can't do that as often here on HN. Here, the quality of article is often higher and the comments are usually more on-topic and, often, more specialized. I do end up reading many more threads than I comment in. But, yes... I learn a lot more from the comments (frequently) than I ever do from the article. HN is particularly good at this, and it's one of the reasons I visit.
- dionidium 9y agoWe're not saying different things. As I said: I love the comments here (usually). I also learn a lot from them. I think "forums" like reddit and HN are fundamentally different from inline comments on news websites.
- KGIII 9y agoYes, I was agreeing with you and sharing my experience. That's why I said similar. Here, I am more inclined to read the article. HN people are just too smart for me to always be able to grasp it from comment context.
- ChuckMcM 9y agoI wonder if there is an algorithm to score comments vs the article so that you can warp the article toward something the commentator would approve. We have already seen examples where people run the same story on extreme left and extreme right web sites to get clicks and rage views and their toning, and changes are fairly algorithmic. That makes me wonder if you could build an echo chamber of one. So that what you read was always something you really agreed with, even if it wasn't written that way for anyone else. (same data and facts mind you, just slanting it toward the reader's preferences).
- seanwilson 9y ago> That's funny. I tend to skip the article and go into the comments. > Comments tell me "what's wrong with the article", and in doing so, give you a good picture of what the article is about. I'm similar. You could think of articles as one giant comment from a single person. I'd rather read comments from several people as it's much easier to spot bias and get a variety of viewpoints.