[barbecue_sauce]

My process with HN articles is

1. Find a topic that looks interesting or important based on the title

2. Read the top two comments.

3a. If they are incredibly polarized, read the article.

3b. If they are in agreement, just read the comments.

[whenchamenia]

That is a recipe for a powerful bubble. HN, while being above average, collectively gets many, many things wrong.

[panarky]

Conventional wisdom says filter bubbles are bad because they hide alternative viewpoints.

But after I've understood and incorporated alternate views and I'm ready to build, then I want to discuss with like-minded people to refine the ideas and take action.

When I'm done with discovery and I'm ready to act, I don't want to keep fighting trolls and pedants, re-litigating stale arguments, weak signal lost in the noise.

More bubble, please.

[la_barba]

Well a bubble is an imagined abstraction. Like most abstractions, it doesn't actually exist. Its just a way to simplify things. Also, there isn't a clear definition of it either, making things even more muddy.

>HN, while being above average, collectively gets many, many things wrong.

That is to be expected. I mean we have programmers opining on quantum physics, bio-chemistry, economics, sociology, and whatever else hits the front page. I don't know of an online community that restricts their opinions to their own domains.

[EForEndeavour]

> Well a bubble is an imagined abstraction. Like most abstractions, it doesn't actually exist. Its just a way to simplify things. Also, there isn't a clear definition of it either, making things even more muddy.

The fact that humans came up with a fuzzy concept to mentally model real-life effects doesn't diminish the existence of those effects. Events occur independently of what names and categories we give them.

For example, hordes of Twitter users have self-selected who they follow until they're exposed to opinions that agree with their own. This is and has been happening out there, in the real world, to real people. Humans, being pattern recognizers, naturally think of this effect in terms of "bubbles" or "echo chambers." These are not defined in a strict mathematical sense, and they don't actually exist, but the phenomenon certainly does.

[la_barba]

>These are not defined in a strict mathematical sense, and they don't actually exist, but the phenomenon certainly does.

That's my point, I'm disputing the central assertion that these phenomenon actually model reality.

>For example, hordes of Twitter users have self-selected who they follow until they're exposed to opinions that agree with their own. This is and has been happening out there, in the real world, to real people.

You can always point to people who unfollow others when exposed to world-view-contradicting opinions. But it hasn't been shown that they always do it, or do it for the same reason in every case. people can and do exhibit contradictory behavior, or have the wrong information, or are simply enjoying the drama, or are trolling, etc. The existence of these and other factors, which are just observations (not models) of human behavior, go against the nice and easy explanations of someone being in a bubble.

We like to put people inside neat little boxes with labels on them. Like you said - " Events occur independently of what names and categories we give them.".

[EForEndeavour]

> But it hasn't been shown that they always do it, or do it for the same reason in every case.

We don't need them to always do it, or for any reason. The fact is that it does happen, as evidenced by, e.g., https://www.cnet.com/news/twitter-echo-chambers-are-very-rea...

[la_barba]

Shockingly, I read the actual study.

"However, as we focused on politically-savvy users on Twitter, the reader should not infer that our observations generalize immediately to other settings, or that echo chamber effects are as pronounced for the general public."

Exactly what I said - " But it hasn't been shown that they always do it, or do it for the same reason in every case."

I don't know why you are pointing to a study that agrees with me.