I tend to be a "comments first to see if the source may be interesting" type; this has some subtle effects on my heuristics. 1) I tend to avoid (the more obvio

I tend to be a "comments first to see if the source may be interesting" type; this has some subtle effects on my heuristics. 1) I tend to avoid (the more obvious) clickbait. Useless information aside it usually spares me the inevitable thread about it being clickbait. 2) I tend to avoid political links. The source is usually politically bent, the comments are polarizing, and it's often unnecessarily toxic. 3) I tend to avoid failure/scandal articles. They are usually just people acting smug because it was obvious that reboxing pizzas at cost was not a sustainable business model. No substance. 4) I tend to enjoy reading technical articles on languages/tools that I use or aspire to use. They're often full of very interesting discussions. I especially like release candidates/patch notes as people often have uses for new features that I hadn't quite considered or groked which makes the value-add all the greater for me. 5) I like a lot of articles about potential futures; where we're going with DNA, with space, with physics, etc. These are things well at the periphery of my own knowledge but seeing passionate people talk about what the future could hold excites me for what I have to look forward to in my life. 6) Beyond those heuristics I also often do a lot of thread-collapsing when I see things start to devolve into zealous/crude mentality. That being said, I often do follow witty retorts and quips, so there's that. 7) I also check out a lot of new languages and tools, not particularly because I want to use them but because I want to know the kinds of problems they perceive with existing tools and how they went about determining solutions to them. It's fun to see some of them start to mature and turn into real things, though (like Rust, which I do explore). As far as actually clicking on the link goes, I usually click on anything that has a few decent comments (or, occasionally, very few comments) and will often read it if it seems highly interesting or opens well and is long-form.