Live data from Hacker News

The Social Graph is Neither

blog.pinboard.in

31–40 of 61 posts

Re: The Social Graph is Neither

#31
post #12

Earlier quoted context omitted.

> Take the most popular person in high school and get him to design social software. It would suck. Bad. Let's assume they had the technical skills to do it, why do you think it would suck?

Let's assume they had the technical skills to do it So basically a nerd with social skills?

Please. For the purposes of this conversation a "nerd" means what the OP wants it to mean, which is at least "someone who knows what a 'graph' is".

Whereas the technical skill required to build a basic electronic social network is "can you dial a phone?" Been that way for sixty years and more. For a dramatic musical rendition of the social networks designed by your parents (oh, wait, I mean your grandparents, maybe even your great-grandparents, how time flies) watch the first ten minutes of Bye Bye Birdie. Ooh, look, a video:

http://www.youtube.com/watch?v=gKhR8QtQ4do

(I was never a fan of Bye Bye Birdie, by the way, but you can't forget that number once you hear the darn thing.)

Now, these days every eight-year-old has mastered a far greater technical challenge: Mobile text messaging. That's a better social technology because you don't have to talk, or say "hello" or "goodbye", or avoid communicating in seventeen directions at once.

There are things Twitter and Facebook can do that can't be done with chains or trees built on one-to-one texting, but I'm not sure any of them are more social. (Is anything more social than the game of phoning people and asking them to relay messages?) Really, the point of these higher-tech "social networks" is that they fill a gap by letting you be barely social and yet still be present in a bunch of people's lives. I can follow your tweets without you even knowing. I can add Facebook "friends" and then basically ignore them most of the time. I can be "introduced" to a friend-of-a-friend on Facebook without any of the real stress of even talking to them, let alone meeting them in person.

Re: The Social Graph is Neither

#32
post #10

This is a superb post. Social networks, starting especially from the era of the Facebook era, have foisted increasing amounts of doublethink on their users, and it's often difficult to make people realize this and explain the reasons for it being that way. This essay not only lays out good examples of that nagging cognitive dissonance, but then goes on to explain its necessity at both the technical and business level…

I disagree. The entire first point is merely a taxonomic complaint. Yeah, taxonomies are hard to "get right" in the global sense. That's why we don't employ globally correct architecture-astronaut taxonomies for everything, there aren't very many accurate useful ones.

Instead life is made up of piles and piles of little taxonomies used for specific purposes. I'm happy with that as both a mental model and a practical one.

I agree that the notion of a single social graph is an incoherent one, but Maciej's point is largely a quibbly one. There are graphs of human interactions that can be mapped out. These interaction graphs are not static. Data sets which purport to represent reality require updating, as they are not the canonical source.

But so what? That's true for maps of physical locations too, or article quality on wikipedia. Facebook shows, if nothing else, that it's possible to not just track but mediate (and still track) these interaction spaces for users as an end to itself (even if they are the product not the customers).

Does it matter that Facebrick isn't a perfectly synced representation of reality? Well, does it matter that Wikipedia isn't a totally accurate encyclopedia? You'd be right to ask for what cause, and to what end.

And that's the whole point. We dig up or create the little taxonomies that bridge the gaps that we're looking to cross.

Re: The Social Graph is Neither

#33

Earlier quoted context omitted.

Exactly. "Social networks" are thingies that aid us with establishing and keeping up with actual social networks. It's a bit like calling a publisher's press a news agency.

In vipassana meditation there is the idea that if you learn to recognize thoughts at the precise moment they arise, then eventually any given thought will basically lose its power and stop arising. Facebook seems to have the same property, but for personality. Every time you post about yourself, you become a little bit more bland and boring of a person. The way I think of Facebook is rolodex + fax machine + succubus.

Fascinating tidbit on meditation, but I disagree that posting about oneself on Facebook makes one more bland and boring. If, for example, one is bland and boring to start with, then yes, posting about yourself to Facebook won't really help. However (EDIT) -- if your life is boring, you can post about something else. I've known some individuals who lead the most pathetic lives you can imagine yet they had a decent-size following on Facebook because they regularly posted about politics (which I guess they used as a medium through which to vent about how pathetic their lives were...).

I do see some truth however to your comparison with that idea from meditation. For example, even if you are leading an interesting life, there are many ways to post about it that will make your life appear less interesting than it really is.

Re: The Social Graph is Neither

#34
post #29

Lots of good, interesting ideas in that post. A few comments: In general, what data you capture, always depends on what you want to do with the data. Any data model is an abstraction, which discards a lot of information. This simplification is what makes the data model useful. Of course, storing social relations as an undirected, graph with a single type of edge, as is done in Facebook, is a gross simplification of r…

Facebook has a heck of a lot more info on relationships than an undirected graph with a single edge type. In the front end alone, they now have a directed "follows" relationship. In their backend, they're keeping track of how much you pay attention to people, comment on their posts, click their links, etcetera, and using it, among other things, to decide what shows up in your news feed and what is silently dropped. Someone posted a link here not too long ago describing how to get the weights Facebook ascribes to your social connections, based on the above sort of metric.

Re: The Social Graph is Neither

#35
post #12
post #8

Earlier quoted context omitted.

This is bullshit. "Computer nerds" have plenty of friends. I know factory workers that have very high depression levels due to not having enough social contact or friendship in their lives. Sure, when I was 9 I was beat the fuck up over and over but when you are friendly, smart and successful people like you. They want to be your friend. Let's play devils advocate for a second. Take the most popular person in high sc…

> Take the most popular person in high school and get him to design social software. It would suck. Bad. Let's assume they had the technical skills to do it, why do you think it would suck?

Let's assume they just have to describe in it sufficient detail to someone with technical skills who would implement it for them.

I'm pretty sure it would still suck because they wouldn't know to translate their intuitive understanding of how to leverage social dynamics to gain popularity with how to facilitate them in software. But it's hard to be sure, and would depend on the latent design skill hidden in your "popular person".

I think most people don't know how to approach software design, even from a UI perspective. It's hard even for professionals. You would have to get lucky with a person who happens to get it intuitively.

Re: The Social Graph is Neither

#36
post #10

This is a superb post. Social networks, starting especially from the era of the Facebook era, have foisted increasing amounts of doublethink on their users, and it's often difficult to make people realize this and explain the reasons for it being that way. This essay not only lays out good examples of that nagging cognitive dissonance, but then goes on to explain its necessity at both the technical and business level…

I disagree. The entire first point is merely a taxonomic complaint. Yeah, taxonomies are hard to "get right" in the global sense. That's why we don't employ globally correct architecture-astronaut taxonomies for everything, there aren't very many accurate useful ones. Instead life is made up of piles and piles of little taxonomies used for specific purposes. I'm happy with that as both a mental model and a practical…

I think you're missing his bigger insight, that is, building a social graph is a fundamentally asocial act. And that extends to creating taxonomies, even if they are localized to my concept of socializing. Nobody goes to a party to create a taxonomy.

Being social is all about disorganization, spontaneity, creativity, breaking out of all the stupid taxonomies that we impose on other people and have imposed on us. Trying to organize socialization to the extent that a "social graph" can be created is a profoundly sociopathic act. To push this task on users so that they do all the shoeboxing and tagging and linking, all in the name of "fun," requires some coercive force. It is no surprise that many Facebook users say that they don't really enjoy using it, but they have to. Part of that was the site's smart choice to maintain real names, so that you are always fretting about your online appearance. It's not a "perfectly synced representation of reality," but it is presented as such, almost like your own personal Wikipedia page, and by the laws of social anxiety you are compelled to check up on it! Shoebox, tag, link...

The line about Facebook rarely producing creative or original material really hit home for me. 4chan and SA are creative because they are fundamentally disorganized, and this freedom produces real moments of brilliance (and plenty more of profound stupidity). Facebook and its ilk are like corporate jobs disguised as consumer web products. You are constantly being shuffled from box to box, task to task, and the blinders have to be tightened ever so constantly to keep you from realizing why what you are doing really never will be "fun" in the way that a raucous party is fun.

About those blinders: those are in a big way the "privacy options" that every social network touts. Maciej points out, and I think that it's a salient point, that if everybody was able to see the whole social graph, it would dissolve. That alone should suggest that the social graph is an unnatural concept. Yet there is somebody, the owner of the social network, that can indeed see the entire graph, and there is a whole level of misdirection that they must apply to make this seem completely normal.

Again, maciej said most of this better than I could, so I'm glad he did.

Re: The Social Graph is Neither

#37
post #8
post #4

> Asking computer nerds to design social software is a little bit like hiring a Mormon bartender. exactly.

This is bullshit. "Computer nerds" have plenty of friends. I know factory workers that have very high depression levels due to not having enough social contact or friendship in their lives. Sure, when I was 9 I was beat the fuck up over and over but when you are friendly, smart and successful people like you. They want to be your friend. Let's play devils advocate for a second. Take the most popular person in high sc…

Take the most popular person in high school and get him to design social software. It would suck. Bad.

What you've just described is Facebook, except that the "most popular person in high school" got cut out of the loop by the nerd smart enough to realize that the idea was good enough to be worth stealing.

And you're right. It does suck, in a million ways, reflecting our society at large rather well, overall. But it's also the most successful social network to date.

But in my opinion, at least, the main idea that the yacht club twins brought to the mix was actually the primary reason that it grew so easily: Facebook was, initially, at least, a social network for students at elite schools, let the rest of the University-of-CrappyPartySchool rabble diddle themselves on Myspace. Without that, I don't know that it ever would have taken off the way it did...

Re: The Social Graph is Neither

#38
""" Right now the social networking sites occupy a similar position to CompuServe, Prodigy, or AOL in the mid 90's. At that time each company was trying to figure out how to become a mass-market gateway to the Internet. Looking back now, their early attempts look ridiculous and doomed to failure, for we have seen the Web, and we have tasted of the blogroll and the lolcat and found that they were good. """

i've been saying this for ages but this guys puts it so much simpler i'm ashamed.

Re: The Social Graph is Neither

#39
post #36

Earlier quoted context omitted.

I disagree. The entire first point is merely a taxonomic complaint. Yeah, taxonomies are hard to "get right" in the global sense. That's why we don't employ globally correct architecture-astronaut taxonomies for everything, there aren't very many accurate useful ones. Instead life is made up of piles and piles of little taxonomies used for specific purposes. I'm happy with that as both a mental model and a practical…

I think you're missing his bigger insight, that is, building a social graph is a fundamentally asocial act. And that extends to creating taxonomies, even if they are localized to my concept of socializing. Nobody goes to a party to create a taxonomy. Being social is all about disorganization, spontaneity, creativity, breaking out of all the stupid taxonomies that we impose on other people and have imposed on us. Tryi…

> Being social is all about disorganization, spontaneity, creativity, breaking out of all the stupid taxonomies that we impose on other people and have imposed on us.

That is a very euro-centric, post-modernist way of thinking.

What are caste system if not social?

What are far-flung familial relations if not social?

A definition of being "social" which only applies to societies where friendship is paramount, and friends come and go with ease is too limiting and leads you to solutions that can be awkward in other societies.

For example, in my country, we care deeply about family and ancestory--so much so that in every generation one member of the family is nominated to be the "memory keeper" (sorry, there's no translatable word) who keeps track of the geneology of everyone, and is expected at a glance to determine if a stranger is at least vaguely related to us.

Would a free form social graph be useful to us---not really. Our facebook pages are practically 90% family members, 10% people who through marriage will become family members.

Re: The Social Graph is Neither

#40
post #29

Lots of good, interesting ideas in that post. A few comments: In general, what data you capture, always depends on what you want to do with the data. Any data model is an abstraction, which discards a lot of information. This simplification is what makes the data model useful. Of course, storing social relations as an undirected, graph with a single type of edge, as is done in Facebook, is a gross simplification of r…

Thanks for mentioning triadic closure. It's a great example of how useful information can be inferred/suggested based on an incomplete or imprecise model. The sacrifice of simplifying the social graph is what makes the task of constructing this graph online approachable.
Post reply on HN