[spamizbad]

I disagree somewhat. Social media apps are powered by feed algorithms that fall into two camps:

The first camp biases toward sprinkling provocative, highly engaged content in your feed even if it falls outside your network of follows or areas of interest. A sort of “forced discovery”. Elon’s Twitter and YouTube during the 2010s follow this model.

The second camp does the same thing but requires recommended content to track closer to its perception of your interests. TikTok does this exceptionally well, to the point where people often say they feel like their feed is “reading their mind”. Bluesky seems to follow this pattern as well.

The latter is more scalable than the former, but to your point it is an open question how big it scales, and maybe there’s just too many people for either approach to work.

[rescbr]

As far as I understand it, Bluesky's default feed is chronologically sorted posts from who you follow. It is as a dumb pipe as it gets.

[Pufferbo]

There’s multiple feeds, which is one of its main features. There’s user created feeds, which are just aggregates of tags and keywords. These are alright, but prone to issues with cross domain terminology. ie, say you want a football/soccer feed and use the word “spurs”, you’ll end up having it filled with basketball and rodeo posts.

The two default feeds are your followed accounts in chronological order, the other is an algorithmic feed. The algorithmic feed is pretty good to be honest. I “disliked” around 20 political posts the first day, and it has seemed to responded fairly quick to that feedback.

[empthought]

No matter how many times I choose "Show less like this" on furry manga art in the algorithmic feed, I still see it show up. :(

[Onavo]

The second camp is just artificially creating echo chambers, a virtual "separate and distinct network" for the parts that matter.

[the_snooze]

The "echo chamber" argument really doesn't speak to me because all I want is a place where I can get timely updates about: people in my research field, pictures of cute dogs, and municipal government activities. The more a website stays laser-focused on my interests, the better.

[cosmic_cheese]

I think those things you’ve mentioned are what most people came to social media for originally, but it’s gotten lost in all the noise. The original point of social networks was to be social and connect with people in your industry or who share your interests or share a locality in common, and X especially has drifted far from that ideal — lots of users now log onto it to find something/somebody to be angry at and to argue/troll.

[mingus88]

I’m still using an RSS reader for that very purpose. I want my trusted content displayed chronologically; miss me with the algorithm and the recommended influencers. I’ve been on the internet long enough to know what I want and how to find it.

[AlienRobot]

I share the sentiment.

It seems some people got on the internet to argue and never learned to enjoy anything. I'm a bit sorry for them, to be honest.

[hooverd]

It's kind of alarming, not that I'm not guilty of it, but you do see people whose entire online presence is just stuffing their face with negative interactions.

[hooverd]

HackerNews is an echo chamber. You're not even allowed to call people f*ggots or speculate about the Jews here!

[derbOac]

That makes a lot of assumptions about the nature of the content provided by the recommendation algorithms, as well as human nature. A good recommendation engine, for example, would recognize when someone either likes a broad range of sources politically speaking, or likes a neutral region.

Conversely, it's unclear that a recommendation engine would be able to predict what would be best at "disrupting an echo chamber", and more importantly, when that is desirable, and what "desirable" even means. It's also unclear that the first model is successful at all in disrupting echo chambers, as opposed to exacerbating or amplifying existing positions. I think there's good reasons to think that provocative can be less effective if anything.

[spamizbad]

I disagree with this: if the only thing you allow to pierce the veil is selected based on engagement metrics you just walk away with a shallow view of your opposition. If anything this may entrench your existing views and give you a false sense of intellectual and/or moral superiority.

You need to “meet people where they are” and the first type of algorithm just doesn’t do that. It just says “conservatives/liberals really like this, so you’re going to be forced to see this too because you show interest in politics”

To give an example: let’s say I’m a small business owner who voted Trump but has some lingering concerns around how tariffs might impact my business. Am I going to be better informed reading some engagement-bait post from liberals talking about how I’m going to get “deservedly” crushed by tariffs or a post from a conservative economist laying out the cold hard facts (both good and bad)?

[Onavo]

Your argument is in support of mine. Separate networks are an interesting legal and software engineering detail, but from the POV of the user, as long as they see what they want to see, they will stay with the network.

[spamizbad]

Ah you’re right.