> "Despite minor differences between individual surveys, the data consistently show that the average number of close friendships rose from 2.2 in 2000 to 4.1 in 2024," says Hofer.
If true, this is an astonishing social transformation, and because goes against everything we here about the loneliness epidemic getting worse.
Or have people redefined what they consider to be "close friends"? Or are people actually genuinely maintaining more friendships because phones make it so much easier to message?
If I'd guess I'd say close friendships meaning is now more shallow. Or: younger demographics are against the wider trend.
We can also extrapolate this to unrelated topics, like friend groups. Granted, completely unscientific. But if you know two or three different friend groups and have a brain cell or two, you'll notice group-member-patterns. The Joker; the athletic; the geek; etc.. The question I'm trying to get to is: will the search for authenticity in a subgroup of a greater acquaintace group push you toward the fringes?
Keep in mind that it is "average" and it is about close friends.
Anecdotally, the pandemic was the great cutting of weaker ties. I talk to far fewer people than I did pre-pandemic (and most friends report the same), but I speak to those people more often. I can easily see that ending in a way where some 20% find themselves with nobody.
I would say I have 4 close friends. But some 10 weaker ties disappeared from my life. Did those 10 also double down on close friends? Or did perhaps some of them not have enough close friends to do that?
I know this paper isn't about social networks, but we know this, we knew it in the 70s. The only difference is that we continue to ignore and forget it.
In-group dynamics are further ingrained as the group gets bigger. If you have 4 friends in a group, their opinions aren't as strong. If you have 40 friends in a group, not only are their opinions stronger, they'll fight vigorously to defend the group's commonly accepted beliefs. So a growing social circle does reinforce the group dynamic. (this is well established by lots of studies)
But increased polarization around the world isn't because of this. There's the typical environmental factors: an increase in changes (or challenges) to traditional values increases polarization; an influx of migrants increases polarization. But then there's also social media, where mastery of "engagement" by businesses for profit has been adopted by political groups looking to sow division to reap the benefits of polarization (an easier grip on power). The rapid rise of polarization is a combination of both.
It's nothing new of course, political/ideological groups have been doing this forever. We just have far more advanced tools with which to polarize.
Polarization maybe a bit unclear word here. Connectivity creates cohesion, which creates larger creatures. So what we have is, virtual monsters roaming around with huge human groups riding on them. They can organize real protests, polarized opinion and massive impact wherever these monsters go.
You can have 10 "friends". 3 close ones. Anything larger than that and you are way out of your depth and can't possibly maintain those relationships in a meaningful, personal way.
This always seemed intuitively inevitable if you ever played with a graph layout tool like dot or similar kinetic layout engine. With weak connectivity the nodes don't cluster readily, but with more connections they "snap" into rigid subassemblies. It always seemed to me like a bad thing for society but it could well be a case of "old man yells at moon."
Links are "DOI NOT FOUND". Article does not seem to suggest that the study actual found any relationship between the increase in the two things, just that they both happened around the same time.
Unfortunately, even for the most fast-moving journals, that time is typically several hours before the actual articles appear on the journal’s website. So, anyone who’s reading quickly is likely to find that the DOI fails.
But that rule only applies to the fast-moving journals, like Nature and Science. Many other journals can take a few days between when they allow journalists to write about a paper and when it becomes available to the scientific community—PNAS, which is a major source of material for us, falls in that category.
The problem isn't connectivity provided by the Internet or the average number of friends. Those things are good on their own. The problem is centralized moderation in an infinitely connective environment (aka the Internet), which will create intellectually and ideologically homogenous groups that increase in size without limit.
The solution is to ban all server-side ranking, moderation, and filtering mechanisms and replace them with client-side-only solutions, at least for large platforms above a certain user count like X and YouTube. Same thing for search engines and chatbots.
Each person should be able to control what they can post and view online, but not what anyone else posts or views. The norms that we use to moderate physical public spaces must not be applied to online public spaces. Until we discard those norms, people will continue to become increasingly polarized, democracy will continue to decline worldwide, and violent conflicts will continue to increase in frequency and scale.
In practice I don't think this really changes anything at least for moderation. It takes a bunch of time and effort to moderate online communities - under the process outlined by the post you linked most communities are going to have a single effective clientside moderation list you can subscribe to anyway.
Totally unmoderated internet communities would be completely unusable because of spam, and it's also questionable whether you could even stay up with no serverside moderation - you'd have to delete stuff otherwise it just takes one script kiddie with a botnet to flood your disk space with garbage.
(User produced ranking/filtering algos though I can see being viable)
There are multiple providers of Adblock lists. It would be like that, not single-provider.
Regarding banning server-side moderation, we probably can't do it without decentralizing content delivery in a BitTorrent fashion. But even half measures like replacing moderators with client-side filters would be a big improvement.
>The solution is to ban all server-side ranking, moderation, and filtering mechanisms and replace them with client-side-only solutions, at least for large platforms above a certain user count like X and YouTube. Same thing for search engines and chatbots.
This is such a HN response. A HN reader might think it's fun to spend a weekend on writing/testing a ranking algorithm, but not the average person. They're just going to use whatever the platform recommends.
It's also impractical even for tinkerers. YouTube claims to get over 20 million videos uploaded daily and it has well over 10 billion stored videos in its corpus. The metadata alone is tens of terabytes. The usual introduction-to-recommendations approaches out there are going to completely fall over on an item set of this size, even if you have disk space to spare.
Most people will use the default algorithm. A minority will choose a different algorithm.
It's only a partial solution. Really, the correct response is regulatory oversight and taxation on remaining economic rent. They are monopolies, and should be regulated as such.
> "Despite minor differences between individual surveys, the data consistently show that the average number of close friendships rose from 2.2 in 2000 to 4.1 in 2024," says Hofer.
If true, this is an astonishing social transformation, and because goes against everything we here about the loneliness epidemic getting worse.
Or have people redefined what they consider to be "close friends"? Or are people actually genuinely maintaining more friendships because phones make it so much easier to message?
Indeed. Conflicting info.
NOTE: I did NOT read the article.
If I'd guess I'd say close friendships meaning is now more shallow. Or: younger demographics are against the wider trend.
We can also extrapolate this to unrelated topics, like friend groups. Granted, completely unscientific. But if you know two or three different friend groups and have a brain cell or two, you'll notice group-member-patterns. The Joker; the athletic; the geek; etc.. The question I'm trying to get to is: will the search for authenticity in a subgroup of a greater acquaintace group push you toward the fringes?
Keep in mind that it is "average" and it is about close friends.
Anecdotally, the pandemic was the great cutting of weaker ties. I talk to far fewer people than I did pre-pandemic (and most friends report the same), but I speak to those people more often. I can easily see that ending in a way where some 20% find themselves with nobody.
I would say I have 4 close friends. But some 10 weaker ties disappeared from my life. Did those 10 also double down on close friends? Or did perhaps some of them not have enough close friends to do that?
>> goes against everything we here about the loneliness epidemic getting worse
This seems like a hot-take. IMHO one does not and cannot cure loneliness by having more online friends.
"An information flow model for conflict and fission in small groups (1977)" by Wayne W. Zachary [https://www.jstor.org/stable/pdf/3629752.pdf]
I know this paper isn't about social networks, but we know this, we knew it in the 70s. The only difference is that we continue to ignore and forget it.
better connectivity -> people finding better friendship matches -> groups are more homogenous -> more polarization
In-group dynamics are further ingrained as the group gets bigger. If you have 4 friends in a group, their opinions aren't as strong. If you have 40 friends in a group, not only are their opinions stronger, they'll fight vigorously to defend the group's commonly accepted beliefs. So a growing social circle does reinforce the group dynamic. (this is well established by lots of studies)
But increased polarization around the world isn't because of this. There's the typical environmental factors: an increase in changes (or challenges) to traditional values increases polarization; an influx of migrants increases polarization. But then there's also social media, where mastery of "engagement" by businesses for profit has been adopted by political groups looking to sow division to reap the benefits of polarization (an easier grip on power). The rapid rise of polarization is a combination of both.
It's nothing new of course, political/ideological groups have been doing this forever. We just have far more advanced tools with which to polarize.
Polarization maybe a bit unclear word here. Connectivity creates cohesion, which creates larger creatures. So what we have is, virtual monsters roaming around with huge human groups riding on them. They can organize real protests, polarized opinion and massive impact wherever these monsters go.
Monsters is a interesting choice of words. Why call it monsters?
Isn't polarization a good thing? If I was enslaved by tyrants making my life worse everyday, shouldn't I be opposed by their ways?
You can have 10 "friends". 3 close ones. Anything larger than that and you are way out of your depth and can't possibly maintain those relationships in a meaningful, personal way.
group think has always been dangerous, 1984 come to mind
https://en.wikipedia.org/wiki/Dunbar%27s_number
Thanks to David Wong for explaining this in JDATE, calling it the Babel threshold.
This always seemed intuitively inevitable if you ever played with a graph layout tool like dot or similar kinetic layout engine. With weak connectivity the nodes don't cluster readily, but with more connections they "snap" into rigid subassemblies. It always seemed to me like a bad thing for society but it could well be a case of "old man yells at moon."
Links are "DOI NOT FOUND". Article does not seem to suggest that the study actual found any relationship between the increase in the two things, just that they both happened around the same time.
Unfortunately, even for the most fast-moving journals, that time is typically several hours before the actual articles appear on the journal’s website. So, anyone who’s reading quickly is likely to find that the DOI fails.
But that rule only applies to the fast-moving journals, like Nature and Science. Many other journals can take a few days between when they allow journalists to write about a paper and when it becomes available to the scientific community—PNAS, which is a major source of material for us, falls in that category.
https://arstechnica.com/science/2010/03/dois-and-their-disco...
The common demoninator is the rise of social media networks.
The problem isn't connectivity provided by the Internet or the average number of friends. Those things are good on their own. The problem is centralized moderation in an infinitely connective environment (aka the Internet), which will create intellectually and ideologically homogenous groups that increase in size without limit.
For details see: https://news.ycombinator.com/item?id=45515980
The solution is to ban all server-side ranking, moderation, and filtering mechanisms and replace them with client-side-only solutions, at least for large platforms above a certain user count like X and YouTube. Same thing for search engines and chatbots.
Each person should be able to control what they can post and view online, but not what anyone else posts or views. The norms that we use to moderate physical public spaces must not be applied to online public spaces. Until we discard those norms, people will continue to become increasingly polarized, democracy will continue to decline worldwide, and violent conflicts will continue to increase in frequency and scale.
In practice I don't think this really changes anything at least for moderation. It takes a bunch of time and effort to moderate online communities - under the process outlined by the post you linked most communities are going to have a single effective clientside moderation list you can subscribe to anyway.
Totally unmoderated internet communities would be completely unusable because of spam, and it's also questionable whether you could even stay up with no serverside moderation - you'd have to delete stuff otherwise it just takes one script kiddie with a botnet to flood your disk space with garbage.
(User produced ranking/filtering algos though I can see being viable)
There are multiple providers of Adblock lists. It would be like that, not single-provider.
Regarding banning server-side moderation, we probably can't do it without decentralizing content delivery in a BitTorrent fashion. But even half measures like replacing moderators with client-side filters would be a big improvement.
>The solution is to ban all server-side ranking, moderation, and filtering mechanisms and replace them with client-side-only solutions, at least for large platforms above a certain user count like X and YouTube. Same thing for search engines and chatbots.
This is such a HN response. A HN reader might think it's fun to spend a weekend on writing/testing a ranking algorithm, but not the average person. They're just going to use whatever the platform recommends.
It's also impractical even for tinkerers. YouTube claims to get over 20 million videos uploaded daily and it has well over 10 billion stored videos in its corpus. The metadata alone is tens of terabytes. The usual introduction-to-recommendations approaches out there are going to completely fall over on an item set of this size, even if you have disk space to spare.
The server can deliver a sparsely randomly sampled RSS feed of embedding vectors and metadata.
Fetch media after ranking on-device.
Most people will use the default algorithm. A minority will choose a different algorithm.
It's only a partial solution. Really, the correct response is regulatory oversight and taxation on remaining economic rent. They are monopolies, and should be regulated as such.
We need to ban the platform from recommending at all.
It would be like more sophisticated Adblock. There are many providers of Adblock lists, but they can't be provided by the platform itself.