Hacker News new | past | comments | ask | show | jobs | submit login
Hyper Social is Dying (medium.com)
48 points by williamldennis on Aug 12, 2013 | hide | past | web | favorite | 58 comments

Quick story that I promise is actually relevant.

When I was watching Kevin Systrom speak about Instagram at the Castro Theater with Kevin Rose back in May, I distinctly remember one particular question. KR was reading a question from the audience to the two cofounders of Insta:

Kevin Rose (Reading a question from the audience): "Hollerback is being called the New Instagram, have either of you used Hollerback?"

KR looks around and says, "Hey guys, maybe I'm getting too old for this, but I've never heard of this. Have either of you heard of Hollerback?"

Kevin Systrom pipes up, totally confused, "No, I honestly can't say that I have."

KR turns to the audience and says "Does anyone use Hollerback?" Complete silence from the audience, crickets even; not one hand goes up.

KR looks out on the audience and says "Is that the CEO of Hollerback up there asking this question?"

"Yes..." A timid voice is heard to say from the audience.

And that my friends is modern guerrilla marketing and how I've heard of Hollerback when they haven't launched yet. Frankly, I thought it was deplorable at the time but in retrospect it did leave an impact on me (whether it was positive or absurdly negative though, I can't tell). I do find it interesting that reading this article made me recall that anecdote so clearly. Branding in action I suppose...

Hehe, classic. That took some guts to owe up to it and sharp of them to call the guy on it. I don't think I would have caught on that quickly, maybe later at night in bed reviewing the day or so.

Kevin Rose is a smart guy.

As an aside, I had the pleasure to meet him at Disrupt in 2010 right after the launch of Digg v4. It struck me as an incredibly interesting moment in time (to set the stage, Digg had just flattened their site, removed granular categorization and was now pulling in content automatically while Reddit was, antithetically, doing the complete opposite).

I talked with Kevin and it was clear to me that he knew what was at stake and that he was doing his best to deliver what he thought was the right answer. It's sort of like 1984 versus Brave New World. The leader in '84 is sort of a tyrant and we almost universally reject his dictum because of this. In BNW, Huxley's leader genuinely believes that what he's doing is right, and he's not sure, but what he is sure is that he thinks what he's doing is the best way to lead.

This was a profound lesson, for me, about management. Even though Kevin Rose was ultimately mistaken about his hypothesis, he executed with conviction and leadership. On some level, you still have to admire the captain of the Titanic, and that's a sort of backhanded way of explaining why I admire Kevin Rose so much.

To fail, and rise from the ashes, is one of the most difficult things in the world, but I think Kevin has certainly done it.

Great story! Hope you don't mind me stealing that one in case I need to illustrate great pre-launch marketing.

No, it's a great one. Feel free :).

I don't think it made a good impression on me, but it was definitely memorable.

Haha, that is brilliant!

It might be mildly interesting, in a whole different context.

This whole write-up is based on the assumption that everyone has hundreds of "friends" on Facebook, but let me tell you something: if you have hundreds of "friends" on Facebook then you're using it wrong. I have 54 friends on Facebook, I know all of them, if someone asks me to add him and I don't know him/barely know him, I don't fucking add him, it's that simple really.

Besides, even if you're so dumb that you can't stop adding strangers as friends, rest assured that on your feed you'll only see activity from the people with whom you interact the most. So, basically, the whole thing is pointless.

I keep hearing this argument but I think it's impractical.

Because every person has a different idea about when and who it is appropriate to friend on Facebook and no one is really right or wrong, it's just a preference.

However this makes for awkward social situations.

For example your college classmate that you don't really know, adds you on Facebook because hey you are his classmate (or teammate). To him, that relationship is sufficient to justify connecting with someone on Facebook. To you, that may be far from it because you choose to keep ~60 friends but he likes to have a broader network.

This happens in all sorts of different contexts. For example you meet someone at a local community meetup, or you speak at a conference and people that you see there will ask to add you on Facebook.

At that point, how do you gracefully handle the situation? Do you really turn around and say "Uh...yeah thanks...but I only keep 'real' friends on Facebook". Or "Uh...sorry I can't commit to a Facebook friendship right now...you are not good enough of a friend just yet to qualify".

No matter how you put it, you come off as rude and weird and that person will likely not even come close to you any more. But your relationship would otherwise be valuable.

That's one of the many reasons I don't have a Facebook account.

The drama at friending time sounds only perhaps a hundredth as dramatic as I've seen on the other end, the unfriend event. "What, you unfriended me because I posted a bible verse last year, what is wrong with you?" "If you unfriend me, I'm telling my mom, your aunt, to yell at your mom to get her to yell at you! (bonus points for hilarity if mom doesn't even use FB)" Even more hilarious is workplace warfare about one coworker unfriending the another where one of them happens to be a (more or less) protected class, like race or religion or orientation or political activism, probably 2/3 the population fit in here somehow, and they play that card in the ensuing battle which inevitably rolls management into it ("He unfriended me and I'm XYZ so he must have unfriended me because I'm XYZ so you need to fire him for creating an unfriendly workplace for XYZ or I'll go to your bosses boss who happens to be a XYZ") And I'm not making ANY of this up.

I agree with your assessment that each person has a different threshold for calling someone a "friend". But, FB lets you reject a request without providing a reason. I can see in your hypothetical situation that one classmate rejecting the other's request could be awkward if they meetup, in a physical sense, periodically. Otherwise, I think one should just reject and move on.

Personally, I try not to use FB as I do LinkedIn. It's strictly for my personal, not professional, network. With only enough exceptions to count on one hand, I don't accept anyone with whom I haven't spent a significant amount of time.

What we're dealing with in this setting is the social pressure fueled by the faulty expectations of people regarding what's rude and what's not. As I see it, that's not something Facebook can address. Not now, not ever.

Wrong. What we're dealing with is a serious market-failure in a truly good system to keep in touch with the guy you just met at the meetup who you truly want to connect with. He just goes and uses Facebook, and you use nothing. Neither is good. And that's the problem, not the rudeness issue.

And I don't have a FB account at all - when I did have one, I used it similar to the way you described. But statistically you and I are anomalies.

The median friend count in 2011 was ~100 [1]. I'm not sure if any more recent confirmed numbers exist, but various articles from 2012 seem to indicate that's crept up to ~130, as well as an average add of ~7 friends a month.

So perhaps the median user is "so dumb that [they] can't stop adding strangers as friends", but that doesn't change the fact that this is the most common usage pattern.

[1] https://www.facebook.com/notes/facebook-data-team/anatomy-of...

Yes, but that only means that users should be educated not that the service should be fundamentally changed. It's like recalling 100k cars because people can't use the stick shift.

If everybody's using it wrong, then it's broken.

THat's an interesting perspective.

Mine is that if everyone is using it wrong, it's time to revisit my assumptions around or definition of "using it right".

I actually just reviewed my friend list, deleted a few but still have about 150 friends. Mostly family/extended family (in-laws, cousins, etc). I'm a dancer on my free time so I have a group of dancers from over the years. I have another group of people I've met and do keep in touch from: church, school, and D&D. I have a few select from work (past and present). In review, they are people I like to know what's going on, but in some cases would never be able to have face to face interaction. Many are on the other side of the country or, for a few, the world. I don't add strangers or and I'm getting picky about acquaintances. I give this as an example of why your idea that of using it 'wrong' isn't looking at other situations.

But don't you have to add some noise to keep the _NSABOT_ happy?

Frankly I don't get all this paranoia about [put here your favorite government agency] spying. You're talking about criminal activities with your friends and you want a more secure way? there are a few. You're not talking about criminal activities but you're concerned someone might actually use gathered data/metadata to glean information about your private life? Then let me tell you something: first, there no evidence to support anyone has ever acted upon that information, second, the government does not focus on random individuals for the sake of it and third, if you're a person of interest, then you fucking had it coming and is gullible on your part to expect the slightest inkling of privacy on your electronic communications.

You shouldn't have been downvoted. But you should know that if you don't tow the party line on this topic you'll be marginalized no matter how valid, on-topic, or sane your comment is.

You're right though. First off, I hope your parent was joking, because if not, what he said makes no sense and isn't really related to the subject at hand. Second, while the whole domestic spying scandal is serious business, it doesn't have to be injected into every conversation. Third, you're right. There is no evidence to support that anyone has ever targeted a random individual for the sake of it and you probably shouldn't be using something like Facebook if you plan to discuss criminal activity, and yeah, no matter who you are you can't expect an inkling of privacy online. But that's not really specific to the government either. I can spy you, the guy down the street can, it doesn't take the kind of infrastructure the NSA has to spy on someone. Hell, I'd be more paranoid that random people are spying on me than the NSA. The randoms are more likely to target you for no reason while the NSA, from what we can gather, is more than happy to collect your data and never look at it - like ever.

Questioning the amount of paranoia we seem to have and being pro-NSA spying are not the same thing and it seems like a lot of people aren't able to separate out some of the nuance.

Yes, I was joking. And it was mildly amusing to see a rant in response. And speaking of towing the party line, I wonder how many things made it to the front page of hn only to be killed… I've already shared my opinions on this subject in regards to questioning[0], so feel free to look if you want more.

[0] https://news.ycombinator.com/item?id=6135337

I knew you were joking it's just that up until this point I didn't intervene in the whole NSA debate and I was a bit stirred to see the thing creep even in an unrelated post, but hey, it was a joke, I get it ;)

…was a bit stirred to see the thing creep even in an unrelated post

It's good to get it out sometimes ;)

Maybe there should be a internet law about how anything being discussed on the internet dealing with technology will eventually delve into something on/about surveillance :P

Since linguistic pet peeves seem to be okay here, let me just add mine: it's "toeing the line", not "towing the line". Think of a bunch of people lining up (for a military inspection, I think, but it could as easily be for a subway) with their toes on a line, so that they're all together in a row.

Thank you. I've always wondered what the correct spelling was. I've seen both ways here and now I have a firm answer.

Exactly, from the previous comment people might have thought that I heard about the whole NSA scandal and was OK with it, I'm not. I'm just saying that in my opinion the whole paranoia is downright unwarranted. I'd settle for quite worrisome.

First they came for people that were doing big things, and I didn't speak out because I live a small life and hope to keep it that way so as to never be noticed by the Eye of Sauron...

One day, a dialog popped up on Facebook suggesting people to put on my Acquaintances list, which hides most of their posts from my news feed. It improved my life.


Second that. Lists were introduced as a response to G+ circles just as the subscribe function was added as a response to the follower model of Twitter. Everyone expected a response from Facebook and they delivered. Then the features were left languishing and were never updated/enhanced. I don't know what's the rationale behind this, lists are actually a wonderful way to catalogue your "friends" and most people would find them very useful, if they knew they're available. Facebook should really consider to make them a more prominent feature and simplify the whole adding process.

Are you sure that lists were added in response to G+? I remember having lists long before G+ was even a thing. I could be mistaken but I'm pretty sure lists and the ability to share things among certain groups of friends had existed for quite some time but was rarely used.

It might very well be, I'm not an expert, I know for sure that they became a thing around the time that Google introduced G+ and its circles.

I think you just made Facebook useful to me again.

"In fact, I’d contend that the friend and follower models can be elegantly replaced by a frequency and engagement model."

Quite honestly, I think FB (as an example) tries to do this already. When I "Like" friend A's feeds more often than friend B's feeds, I see more of friend A's feeds. As far as "engagement" goes, I assume the author means "quality of engagement". How do you begin to come up with a way of quantifying the quality of an engagement for one person versus another? Time of engagement? Number of words spoken/written?

"Based on current smartphone technology, social networks can and should leverage location, time, frequency of interaction, and behavioral similarities."

Smartphone, laptop, whatever, they're just data collection vehicles. The difficulty is developing a model that maps values to people. Someone optimizing a social network for behavioral similarities, for example, fails if the individuals in the network thrive on behavioral differences.

There is plenty of room for improvement with social networks. I don't think the way forward is for social networks to improve on network building optimizations.

I could have expanded on the frequency and engagement point. Currently, the most representative social interactions happen off of services like FB (such as texting, phone calls, going to the movies, grabbing coffee, meeting someone, etc).

FB then relies on the user to input this data with the promise that the data will be spit back out in a relevant, useful, and entertaining way.

Phones, contrastingly, can pick up all this external engagement passively and produce a more representative of a true life social network. Because its data collection is passive, it can dynamically adapt a network without explicit user input.

> The central mechanic and experience of these networks, the feed, becomes worse as the network connections increase. This is a huge problem. As more friends enter your new feed the strength and relevance of relationships decreases. The experience should be getting better the more you use it.

Thus why lists were created. Google "got this right" with Circles.

> The next great social network will provide its users with an experience that values friendships and relationships dynamically.

Basically an "Auto-Signal-Noise" generator. I get it, but I don't see it happening anytime soon.

The problem is most people can't even assign true value to their own friendships. Why is a piece of technology going to do that any better?

I wrote about my thoughts more in depth about this awhile ago: http://www.techdisruptive.com/2012/09/18/we-are-far-from-sol...

"Based on current smartphone technology, social networks can and should leverage location, time, frequency of interaction, and behavioral similarities."

And this sort of approach encourages a power-law distribution in who ends up in my feed and that's misguided.

The thing is, actually, I want to see the important updates from all of my friends on FB and the fluffy stuff from only my closest friends and family.

It's not just who, but what -- and yes, that's harder to algorithmically assess, for now, but that doesn't change that anything short of it isn't the big-win solution as the author proposes, IMHO.

I definitely see the value in receiving these kinds of wide-reach-yet-important social updates.

However, I think the opportunity (from both a business and user experience perspective) is for a true digital representation of your current real life social network.

Except, I reject that there's a difference between my online life and my "real" life. You're thinking 20th century/pre-mostly always connected world.

I care deeply about friends and family that I cannot see on a daily basis because of distance and time. That doesn't make those bonds any less real or important and it doesn't mean that online socialization (jokes aside) is any less meaningful than "real life" (there are, in fact, scholarly research publications that support this and I could find them, but honestly, it's been a long day).

Moreover, frequency of interaction doesn't always reflect strength of social bond. I have friends I would kill for who I talk to once a month or less, simply due to schedule, new parenthood, etc. I think that's certainly the case for many adults.

Keep in mind that the dynamic argument also applies to the information that you give out and the content that you create which I think is even more important.

For example, in traditional relationships, the closer you are to someone the more they know about you.

Your parents probably know many things about you that your wife/husband and your best friends don't know.

Your wife/husband knows things about you that your best friends don't know.

Your best friend knows things about you that your friends don't.

Your friends know things about you that I don't know.

And it goes on and on...

This feature of traditional relationships is extremely important and valuable to me.

If you get out of touch with someone, their knowledge of you gets outdated. As you become more intimate with someone, they learn more and more about you.

So there's a natural self-adjusting and dynamic layer and hierarchy to traditional relationships that keeps everyone sane and allows for longevity of relationships.

Social networks throw everyone in the same bucket. Granted they provide you with tools to try and manage the mess like the circles and lists. But that is not really dynamic and to me is very unattractive, imperfect and time consuming. I'm not willing to spend hours managing my social network.

Traditional relationships take care of themselves.

As an Australian living in the US I've come to rely on facebook far more than I ever did when I was living back home. It really is the main way I keep in touch with my friends and family back home. I have a small circle of friends who I chat with on IRC every couple of days, but the time difference generally results in this being a "ships passing in the night" type catch up. My nuclear family I have probably monthly google video chats with. I swap emails with my Dad once a week or so.

I check facebook daily and really feel like it has made it possible to still be part of my friends' lives.

I have run into a bit of a niche use case issue recently though. My hobby is skydiving and facebook has become the way that I get access to photos and videos taken by my new "skydiving friends". So it is quite a regular scene that after jumping with a group of people for a day we'll all trade facebook details, add each other as friends and tag everyone in photos and videos.

This is excellent.... except that these new found friends have access to all my facebook history unless I work out some way to manage the privacy features.

I looked into this today and the "acquaintances" idea kind of fits... except this lowers the weight of these people's updates making it into my news feed, which isn't what I want. I'm really keen to see skydiving related posts from this group of people. I want to know the next time they're planning on a trip to the dropzone. I want to see their other skydiving photos.

So I make a skydivers group and I can limit my old posts to not show up for these people. When I make new posts I can include them if it is skydiving related. But I'm relying on them doing the same. I can't see anyway that facebook makes it easy for me to filter their content. It seems all or nothing. Either I get all their updates or none.

So, basically, the solution to "hyper social" is even more social data, automatically gathered and funneled to company X, so it can be more context-sensitive in what it shows you.

Forgive me for not leaping for joy at the prospect of this future. Imagine if FourSquare were always (literally) on, and tracking literally everyone you meet. And then they started selling your data for advertising, as they're doing now. Or they get hacked, and someone dumps the data / performs enough queries to make an extremely detailed ID-theft of a handful of random people.

I'd rather not have all my past, present, and future eggs in the hands of one anything, much less a for-profit company.

The value and utility of such a service will be so significant that it will trump most individual's concerns about privacy. Facebook has lead a successful charge in this direction for the last 9 years.

I'm not sure that dynamic segregation of social connections is the ultimate solution to the mediocrity of social network applications. Or rather, most social network sites come to serve a specific purpose or narrow purpose rather than actually facilitate social interaction.

Facebook, LinkedIn, and Twitter are just as the author describes, "a digital filing cabinet that you fill with your relationships." They have peripheral uses like being a channel for recruiters and being quoted on mass media, but fundamentally they store contact information.

A "digital filing cabinet" for storing slightly different sets of relationships seems like a tool with limited use, rather than a true virtualization or revolution of social interaction.

I would say that social networks or "social spaces/contexts" (maybe someone can coin a better term than me) that actually facilitate meaningfully novel kinds of social interaction, start relationships, and generate content are more along the lines of HN, reddit, IRC, and forums (online spaces where ideas are exchanged, discussions and learning occurs, and people connect to new people).

Others include Meetup.com, which focuses on connecting people with similar interests so they can interact in real life.

Surprisingly I am more social on Steam than any of the conventional social networking sites, Steam is a surprisingly effective social network for gamers. Additionally, social games like MMOs or other multiplayer games change the way people interact socially through guilds, alliances, clans, trading, competition, and cooperation.

Generalized social "filing cabinets" like FB, LI, and Twitter serve their purposes and being generalized, they have use for a large population of people, but they don't actually generate or enhance interesting new social dynamics or content (twitter sort of does). Friendships and relationships are difficult to cultivate without specified context. A "filing cabinet" has little context for interaction, but a "social space" can provide context for new interaction and that's where more interesting stuff happens.

I don't really like Facebook, but those are non-problems really. I have about 250 connections, which I guess is pretty average or to the lower end. In the past, I would actively go through the list and purge people who I only met a few times or fell out of contact with, but at this point if I see something on my feed that is not interesting to me (either a person I'm not interested in anymore or someone who posts a lot of annoying stuff) then I just hide all their posts. That made my news feed very clean. It also safeguards against the possibility that I might run into one of those people again, it would be very awkward if I deleted them completely.

The social networks of the future will have the potential to access all of your daily interactions via your smartphone. They have the potential to know who you call, who you meet with, and where you hang out.

This enables your social network app to keep up with your relationships, without you needing to tell it that X is now just an acquaintance and Y is suddenly an important person in your life. In fact, your app might figure those changes out before you do.

Even under the threat of panopticon, these advancements are attractive.

Why couldn't facebook adopt this dynamism? I expect they do some of this already, in putting more items in our feed from people we're engaged with. Why not incorporate geographic proximity too? It's feasible (if scary) to bump posts of people we've called. The data is on the phone, no?

I don't see why a new service has to be created for this, when Facebook already has a tweakable algorithm.

Part of the main problem I see if the "friending" model. There's too much friction there to maintain a truly current social network.

I don't even think of Twitter as a social network anymore (probably never did actually). To me it's a hodgepodge of really nichey news, occasional customer service interactions, random personal asides that I don't care about but are mostly harmless, with the occasional personal interaction sprinkled through out.

Some of the younger people I know use it as a chatroom -- they tag messages with usernames to draw people into discussion, and the frequency of messages is usually high.

I do not understand how they are comfortable with having what often amounts to quite private conversations out in public. I found I had to unfollow most of them, the constant and mostly irrelevant noise ends up polluting my feed.

Facebook already does this by showing you only a part of your friend's information on the newsfeed, based in part at least on interaction type and regularity. Google does this via circles. I believe twitter is not curated in that fashion but twitter has a slightly different model due to it's type of asymmetry.

The frequency and engagement model seems to have merit. I might add to the other concept described -- Core vs Circumstantial Relationships -- a notion of Aspirational -- under which following a celebrity might fall.

These arguments are about the same category as stating that email is broken - If you have 600+ friends or emails that overwhelm you, blame your process and not solely the tool.

Facebook does make this dynamic and based on engagement. If you look at photos of someone or send them a message their relative frequency in you feed goes up...

"The Next Social Framework Will Be Dynamic" - I think this is what Google is trying to do with "circles".

Maybe - but at the moment they just make the problem worse, because you have to manually curate who's in which circle and make sure you keep it up to date.

One giant argument from anecdote.

lol, exactly.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact