GeekBox

SF Chronicle Summarizes SNS's

CLICKING FOR CONNECTIONS / After big splash, social networking sites try to hold onto users.

Orc Hut?

So Google is apparently getting into the Friendster business. Along with all the other social networking software that is out there they think there is room to add another. The beta is called Orkut after the Stanford grad and current Google employee that started the project.

Far be it for me to make fun of someone else's name since I endured much teasing as a child for mine, but Orkut just doesn't roll off the tongue in English as a pleasant cozy sounding place to meet new friends. If this is indeed a serious project at Google I strongly recommend that they rename the thing.

Google is a quirky, geeky name, but it somehow worked as a brand and is so popular that it is in danger of becoming a generic term. Somehow, I just can't imagine people saying "Hey Orkut me a message dude", or "She's my best Orkutster".

Tribe Tips

With anything that grows at an exponential rate, there is a tipping point at which it takes off. Tribe.net hasn't quite gotten there yet, but I'm starting to see signs of something new and different. As I was trying to point out before, it comes down to subtleties.

In this particular case, I suddenly noticed the utility and novelty of the simple fact that you can see to which tribes other people belong. It may not seem like much, but it's actually a big deal. Part of what made Napster (and its successors) useful as a means of discovering new music was the ability to browse other people's music collections. Chances are, if someone's collection contained music that you really really liked and some that you'd never heard of before, the unknown stuff was worth checking out. (c.f. my post on Echocloud)

People on Tribe generally don't list enough musicians, authors, or films to make this sort of cross-pollination work as effectively as it does on a P2P network, but they do list every public tribe to which they belong. Tribes are growing in all different directions, with little centralized planning. The directory is becoming unwieldy, which at first seemed really irritating to me. However, if you poke around your friends and their friends' profiles, chances are you will discover some interesting sounding new tribes amongst ones you're familiar with.

I'm sure that there must be some prior art out there - after all people have been attacking the mailing list/newsgroup/forum/chat room taxonomy issue for a while - but it seems like Tribe has hit on one of the things that will make it special.

Back when I was in college, it was often possible to look at other people's .newsrc file which was a list of the Usenet newsgroups to which they subscribed. At the time, I think that many people would have been horrified if they realized that others could see their subscription to alt.sex.bondage (but then again I wasn't living in San Francisco). It was an unintentional byproduct of a wide-area file system (AFS) and lax unix permissions. It also wasn't useful in the way that Tribe's public group memberships are because a) you often had little idea whose account you were snooping around because you didn't have much context and b) Usenet had a rigidly controlled taxonomy (with the major exception of the alt.* groups) which was a little more self-explanatory.

The free-for-all creation of tribes is sometimes inefficient and duplicitous, but it really doesn't matter when you have your friends' interests to go by. Just like you can't devise a perfect taxonomy for music, you can't devise a perfect taxonomy for social groups or interests, especially when you're dealing with a lot of people who are trying very hard to not be mainstream.

The other part of this equation that I'm looking forward to is the "events" section. Many years ago, Craigslist's events section was a good source of interesting things to do, but it's gotten so big and general that it is too hard to sort the signal from the noise. I've also been subscribed to SFRaves, the Squidlist and Flavorpill. I have found good stuff through these in the past, but frankly I see more promise in the events part of Tribe. In a place like San Francisco where there are so many things to do, option paralysis often sets in.

On Tribe, individual tribe members can post events. Somehow, having a listing for something from someone that you know, or at least have some connection to, makes the event seem more like a party that you happened to hear about in class in college. It may end up being random, it may not be the optimum choice for the evening, but it has a better chance of being worth-while than something picked out of a list in the newspaper.

If I'm looking for a movie I'll probably stick with Moviefone.com or Yahoo. If I want to see all the bands coming to town in the next week, I'll stick with the SF Bay Guardian. At the same time, the Guardian listings are often overwhelming. Smaller more offbeat things are better handled by something like Tribe. If you're looking for a low-key weekly event at a club or bar, the Guardian or lists like SFRaves are often out of date. They publish the same weeklies all the time and don't have the bandwidth to check up on 100's of events that change all the time even though they're conceived of as ongoing concerns. As long as Tribe events aren't ever set up to repeat, real people will have to post them, and that makes it more likely that the events are real. Since you can see who posted the event and how they're connected to you, you also have more info to go on when planning your social calendar.

I'm not expecting Tribe to become perfectly omniscient but I can definitely see the opportunity to take advantage of your friends and their friends when it comes to finding new groups and things to do.

Six Degrees Away

With any given subject, it is usual that most of the news articles you read will be largely a rehash of ones you've read before. This article isn't one of those, at least for me. While some of it is a rehash about Friendster, Fakesters, and 6 Degrees of Kevin Bacon, the connections to a mathematician named Paul Erdös are new to me, and explained in an entertaining non-academic way.

[via waxy.org]

Friendster Exposed

This is something that I originally wrote and posted on Friendster about the first Friendster Exposed party which was held in San Francisco, June 5, 2003. I just ran across it again and thought that it was worth sharing...

Well, as could be expected the party was tame and laborious. I went hoping that there might be a bunch of people that I would know, who could, in turn, introduce me to the people there that I didn't know. Instead, I didn't really know anyone and I reflected back on the orientation week that I went through as a freshman in college. Thankfully that painful concept passed out of my mind, but the party still wasn't exactly what I hoped for, even in my least wild dreams.

The only people that I recognized were:

1. Jonathan Abrams (the guy that started Friendster)
2. A friend of a friend whom I recognized from his Friendster picture.
3. That guy Frank Chu that walks around the city with the "Impeach..." sign all the time.
4. The doorman who I guess is one of the Friendster employees (Steve is his name).

The club (Cloud 9) is a nice medium sized space. It has three floors: ground, mezzanine, and a basement. Each has a bar, and the main floor and the basement have their own separate sound systems. The music wasn't terrible, but it wasn't overtly good either.

There was some sort of camera crew walking around with professional video and sound equipment recording this event for posterity. I avoided the business end of the camera.

Some random woman walked up to me and said that she and her friends were playing this game, called the "rejection game" where you're supposed to walk up to as many people as possible and see if you can get rejected. It was a clever idea for a pick-up technique, but she was going about it all wrong. She didn't even try to ask me anything that I could reject until I prompted her. I said that assuming this was some sort of reverse psychology idea, wouldn't she at least want to ask for whatever she wanted in case I actually accepted? She then asked me if I wanted to dance - nobody was dancing on that floor and it seemed really out of place. So I rejected her. Maybe this is what she really wanted. I really don't know.

It is rare that a woman hits on me in such an overt fashion (assuming that was what happened). Usually when it happens there is some element that is oddly out of place. In this case, while she was probably older than me, she had braces on her teeth.

The only useful thing that I learned is that I remember people's names much much better when I see a name and a face together. When someone introduces themselves their name usually goes in one ear and out the other.

Putting your name and picture up on Friendster could almost be like this guy Scott that goes around with a name tag on all the time. Maybe there is something to that idea...

A Group Is Its Own Worst Enemy

I'm not even done reading this, but if you have any interest in on line communities, go read this article: Shirky: A Group Is Its Own Worst Enemy.

Friendster vs Emode?

So according to CNET, Jonathan Abrams et al have managed to pick up a little extra spending money. $1,000,000 in angel funds. The obviously good thing is that they can buy more servers. About time.

Buried in the article are a few interesting tidbits: Emode (doesn't that just remind you of "commode"?) is jumping on the bandwagon with the Emode Friend Network. Yawn. Please. Make. Them. Stop.

(I joined Emode to look around, and people there are so fucking normal it's frightening. I think they get them out of stock photo CDs or maybe they grow them in vats. Seriously there aren't that many boring people here in San Francisco, even in the Marina. That's why I live here.)

Even more clever, Friendster is planning to add IM.

Hello! Hey kids, remember last century when every web site had the portal disease? Then everyone flamed out and blew up? Well, looks like no one learned anything. Concentrate on your strengths, don't try to be everything to everyone. You end up pleasing no one, and it sucks down a lot of resources to try to do everything.

[via Anil]

Tribal Futures and Tribal Past

Last Monday I participated in a focus group about Tribe.net. I've been meaning to write something about it since then, but I've been lazy this week.

I signed a piece of paper that said that I wouldn't reveal any personal info about the other people in the focus group, but oddly, they didn't make us sign a non-disclosure agreement. Well at least it seemed odd at the time. Now that it's past and I've read a bit more about Mark Pincus, the CEO, I get the feeling that Tribe is consciously or unconsciously trying not to be that corporate. I'm taking that as a good sign.

There were about 8 of us including danah boyd who lead the discussion. She works for Tribe and she's also a Ph.D. candidate at UC Berkeley, amongst other things. She asked us all to introduce ourselves and tell the everyone else why we were there. I ended up summarizing and referring to what I had written here about Friendster and Tribe.

It would be safe to say that the people there were early adopters, seeing as Tribe had apparently only been live for about two weeks at that point. Still, some people had much more depth of experience with the net than others. I'm guessing that was intentional - I don't think Tribe is just trying to appeal to hardcore geeks. Everyone had used Friendster, and that was often a reference point for our discussions. I'm sure that Tribe would prefer that they would be judged on their own merits and not in opposition to Friendster, but it is inevitable, seeing how popular Friendster is right now. Ryze, Linkedin and Sixdegrees also came up, but not everyone there had even heard of them.

We discussed Fakesters and the quick path that many Friendster fiends have between "wow this is so cool" to "this is so over". Unless Friendster adds new features and cuts down on the draconian obliteration of the Fakesters and real people that are mistaken for Fakesters, it will become like a one hit wonder and fade into the past. Apparently Friendster has recently implemented a 200 person limit on your friends. While that will probably be fine for most people, I have some hyper-social friends that are bumping into this absurd rule already. Friendster seems really out of touch with its populace.

Everyone seemed to like the fact that Tribe adds ... um ... "tribes" to the mix. Tribes are really yet another incantation of newsgroups, listservs, mailing lists, bulletin boards, or Yahoo groups. I'd like to reiterate that the subtleties of how moderation, threads, replies, creating new Tribes, splintering old ones etc. are handled will make or break this feature. A lot of different organizations have re-invented this particular wheel many many times since the beginning of email, and I'm not sure that anyone has ever gotten all the details right in one instantiation. This is one area where I think it would be useful for Tribe to do some serious research and cherry pick from the best of what's come before. This includes custom systems that only are available to students at particular colleges, the Well, eGroups, and the first public computer bulletin board system, the Community Memory Project. Even the punk/emo/goth-porn site Suicide Girls has an interesting bulletin board, group, journal and friend system that's worth investigating.

I also have a fundamental problem with the fact that Tribe is creating yet another parallel universe. It used to be that if I wanted to know everything that was said "publicly" on the net about a given subject, I would just subscribe to the appropriate newsgroup, or perhaps I would have to be on a mailing list as well. Sometimes, people involved in certain topics would set up gateways between those worlds. Subjects were available as newsgroups or mailing lists, and even in daily or monthly digest formats. These days, there are countless web sites each with their own bulletin board systems, blogs with discussions, mailing lists, and gated communities like AOL, Yahoo groups or Tribe. As a media junkie and geek I find it overwhelming. If I take a topic that I love to follow like Apple Macintosh computers, digital cameras, or mobile phones, I end up reading about a dozen sites - for each topic - a day to keep on top of what's going on. This doesn't seem effective or rational. Some may point out that RSS helps re-integrate all of your news sources, but to me it just ruins the experience of reading a story that is on a well designed web site.

Usenet was once a very powerful and amazingly useful system, which is really hard to explain to someone these days. Unless you were talking about an "alt" group, people had to vote to create new groups or reorganize old ones. There were FAQs for most groups, and there were often some very smart people who posted on a regular basis on a given topic. Hell, even alt.sex used to actually be informative and interesting. Unfortunately, spam and the rise of the web killed the relevance of Usenet to most people. Of course Usenet is still there, and Google has done a largely fantastic job of making the past accessible and searchable, which is something that used to be impossible (hence the necessity of FAQs). However, the spammers and the vast unwashed masses have made the signal to noise ratio of most groups near zero. Occasionally I search for something on Usenet using Google and find useful info but almost invariably it is buried in shit. I haven't bothered to read any newsgroups on a regular basis in years.

In a sense the fact that the net keeps dividing itself into groups is understandable. I can't find a good reference right at the moment, but I've read in the past that there are optimum sizes for certain kinds of social groups - I remember that number being around 200. Certainly the most interesting mailing lists or newsgroups that I've been on have a core group of regular posters and then lots of lurkers, with the number of people being somewhere between 50 and 1000. If the number is too big it is hard to have enough shared values to carry on discussions, if the number is too small, there isn't enough diversity to keep things interesting. So it makes some sense that we can't have one central discussion forum for everyone on the net who's interested in a given topic unless that topic is sufficiently esoteric. So the fact that there are multiple places on the net that I could go to discuss Mac OS X, for instance, isn't that surprising. Or the fact that there is only one mailing list that one should join if they are analog synthesizer fanatics makes sense too. (Of course this is ignoring the fact that not everyone in the world speaks the same language.)

In the print world there are numerous magazines that cover similar topics, yet they all have the same interface, and given a sufficiently large store, you can buy them side by side. In some strictly utilitarian sense this is a duplication of effort, yet in another sense it is just a good example of capitalism at work (well at least in theory). On the net there is no guarantee that an individual will easily be able to discover all of the bulletin boards, newsgroups, mailing lists, etc. devoted to a given topic. And each one has its attendant differences in UI and availability. Any given magazine's availability is certainly limited by distribution, but I can read any of them using the exact same methods.

Despite many companies efforts to bastardize and brand email, it is still fairly standard. While some of the formating, attachments, or HTML email (just say no) may get munged going from one system to the next, in general any email client can communicate with any other email client anywhere in the world. One can't say that for bulletin board systems.

After the official part of the focus group was over, Mark came in from behind the one-way mirror to introduce himself and invite us out for a drink and more discussion. From the focus group and the conversation that ensued afterwards it's clear that Tribe is conceived as much much more than just a "Social Networking System". The elevator summary is that Tribe is a cross between Friendster and Craigslist. In some senses that's true - Tribe handles the basic network-of-friends idea of Friendster (minus its dating myopia) and it will be supported by paid "listings" much the way that Craigslist is supported by paid job ads. However, Mark seems to have much more overarching (possibly unbridled?) vision for Tribe. It has the potential to become a way to represent yourself to other people on the net.

On Tribe your immediate friends can see your last name, whereas your second degree only gets your first name and your last initial. This is a really basic example of the sort of intermediation that something like Tribe could do. In your profile, Tribe has tabs for Basics, Personal, Professional. At some point they may figure out ways of sharing your information depending on the context. Even in San Francisco not everyone wants to advertise their fetishes along with their resume. Tribe, like Craigslist and other systems, allows anonymous postings which still let someone reply to the post without revealing your email address. This is a really important feature that works on Craigslist or even Slashdot because of community moderation, something that Friendster is totally missing. It sounds like Tribe hopes to go beyond this binary choice of anonymous or not to allow shades of gray in between. If they can figure out how to make this work, I think that they will have something very compelling.

Given the massive problems that spam causes and many peoples concerns with privacy on the net, I can certainly imagine that there could be ways of using a trusted service as an intermediary - something that can control how you are represented on the net. One of the main problems of course is that unless a system becomes pervasive, it is likely to be a hassle not worth dealing with for most people. Mark was also talking about ways of displaying listings on blogs or other non-Tribe pages. I could see this working something like Google's ads, which many people seem to like much more than the usual banner ads.

I'm sure I'm missing some important threads from our discussions, but in short, I think that Tribe has some good ideas and I'm looking forward to the execution.

Tribe.net Strikes Back

Funny how this intrawebnetsuperhighway thingy works. I posted my Attack of the Clones story about the backlash against Friendster, and then I posted a link to the blog entry on the very bulletin board system on Friendster that I was trashing, and within an hour the very person that invited me into the Friendster fold in the first place had the solution to my problems: Tribe.net.

Tribe.net is yet another virtual community aggregation site on the web. There were others before Friendster, such as Sixdegrees that flamed out, and there are others such as Ryze, LinkedIn, EveryOnesConnected that are around right now but not as cool as Friendster. Tribe.net seems to have learned from these other services. Plus they've paid attention to some of the points that I raised about fostering discussion and community through moderation and linking people that aren't in your immediate network. The mere fact that they have a pop-up menu that shows [anyone/4 degrees/3 degrees/2 degrees/a friend] is a huge improvement over what Friendster is doing. They get it.

After looking around the site for an hour or so, I have to say that it seems like the right direction. I already notice some of the obvious net.personalities such as Justin Hall and Xeni Jardin have picked up on it. In fact if I had been reading Boing Boing more carefully I might have picked up on this a little sooner.

Attack of the Clones

I've been talking to a friend of mine at Salon.com about the whole Friendster phenomena for months. She's been planning to write something about it for a while. By this point there have been tons of articles and blog posts about what Friendster is and why it's so cool, so of course now we're onto the backlash stage. She finally got approval from her editors to write something and her story is up. Co-incidentally, in the course of a couple days, the Village Voice, Salon, and the SF Weekly have all published very similar articles about fakesters and Friendster's reaction to them. It's funny to see how stuff like this goes in cycles, and the same ideas come from multiple places at once.

I've been on Friendster since the middle of March. My userid is around 50,000. Now they're up to over 1.6 million. It's been growing at an insane rate. Within my group of friends I've seen some interesting patterns of usage. Some people politely sign up and maybe even put up a photo and a simple profile but then they hardly use it. Others ignore the first or second invite and think you're a loser for even suggesting that they would want to sign up for something so dorky, but then somehow after 5 friends all invite them they get the idea that it isn't just for nerds or spam magnets. Then they sign up and become completely addicted. One of my friends finally went cold-turkey and took down his profile after logging in and tweaking his profile multiple times a day. There's a well written story (by someone I don't know) that follows the arc of a Friendster addict beautifully.

Meanwhile Jonathan Abrams et al, are trying to monetize eyeballs, but unfortunately they don't seem to have much of a clue about fostering communities. I wasn't a big fan of the fakesters on Friendster, but seeing this really corporate, dull clamp-down on what Friendster's users are doing, I find myself rooting for the fakesters. You can tell that Abrams doesn't get what he's got. It's almost as stupid as the RIAA discovering that a bunch of music-fans/customers are really happy because they are able to find interesting music on P2P networks and then suing them. The only salient point that they see is that people are "stealing" music, while missing the features of the P2P networks that people would gladly pay for if they were done right. Apple at least understood that you should give the customers what they want. If it is cheap enough and instant enough the "free" version which involves stealing becomes less interesting. Of course until they manage to have the same breadth of choice as the P2P networks I don't find it that exciting, but that's another post.

I first experienced online communities in 1986 while interning at IBM. I was able to read and post on their internal forums. At the time there were some uber-geeks that were allowed to access Usenet, but the forums that I was on weren't linked to the outside world. Even at such a large staid corporation, I could see what I would latter come to know as archetypal patterns of mailing lists/Usenet newsgroups/forums/bulletin boards etc. Excessive cross-posting, arguments about the topicality of posts, flame wars, experts, newbies, FAQs - none of that is new. While in college I started and ran a couple of mailing lists: one devoted to DJing and another about "post-modern culture" and art. I've played around with IRC, MUDs, chat rooms, IM etc. More recently I've played around with Nerve, Match.com, eMode, and Alt.com. And of course now you're reading this on a blog which has the potential for feedback and conversation although it is primarily a unidirectional system. At some level all of these are media for meeting other people and talking about shared interests. Each medium has some obvious big differences and not so obvious subtle differences.

In any system there are people that will try to destroy it by posting troll messages, spamming, or otherwise shaking things up in an anti-social manner. Even early in the history of Usenet when you had to work at a big high-tech company or go to a geeky college in order to have access, there were people that had the online equivalent of Tourette's syndrome. In order to have a viable community there have to be rules and ways of enforcing them, otherwise the system falls apart.

Part of Slashdot's success is its implementation of moderating. By giving its users the ability to self-police and rate posts, they've substantially increased the signal to noise ratio of the discussions. Kuro5hin uses a similar but different rating/content system and achieves impressive results. Craig's list is legendary for its sense of community, even though it doesn't really offer much in the way of dialog. What it does do right is let the community filter and moderate the posts, while it intentionally doesn't encourage replies and discussion threads. Each of these systems is slightly different in how it handles anonymity, replies, and ratings, but these are all features that have to be addressed in a consistent way for a system to hang together.

Abrams started Friendster as an online dating service, but he's really missing the point. Friendster is already a virtual community. It has some good features that make it exciting, such as the ability to post more than one picture of yourself, show how you're connected to your friends and acquaintances, and write testimonials about each other. This is the part that was well done that caused the system to grow exponentially. Once people have been on for a while they look around for something else to do besides collect other Friendsters. The fakesters are a reaction to the fact that other features of Friendster, such as the bulletin board system are really broken and lacking. People are looking for a creative outlet to communicate to others on the system. Every viable online community has a decent communications medium. On Friendster the bulletin board system seems to be an after-thought.

When I first joined, Friendster's one bulletin board was open to anyone in your 4-degree network, now it's cut down to your immediate friends. Both systems are broken. When it was open to zillions of people the signal to noise was near zero, and there was no way to carry on an effective thread of a conversation. Now that it is only open to your immediate friends there is no conversation because if person A posts and person B replies the audiences for the posts are different. The first post may only be seen by person A's friends and the second only by B's. Presumably A and B are friends and may have several mutual friends, but it's next to impossible avoid forking the conversation threads. Even if you accept that it is primarily there to post announcements it is too limited. If I wanted to invite people to a party I should be able to at least add the second degree of friends to the intended recipients. There is no provision for this.

Friendster profiles have fields for books, music, movies, tv, and general interests. It even automatically tries to link your entries to a search for people with similar interests in your extended network. This seems like a natural jumping off point for a topical bulletin board system hosted by Friendster. Instead I'm stuck with one option - I can email some random person with a dorky message saying "Gee, I see that you like collecting Pez dispensers too, wanna be my Friendster?" A mailing list or bulletin board can be like an ongoing virtual party. It can be an interesting way to meet new people. If I choose to be semi-public by posting to a board, I expect that others would be able to contact me. In fact I'd be happy if the boards were open to anyone on Friendster that wanted to join that topic. Filtering by who is in my little clique of friends is antithetical to the supposed point of the whole system which is to meet new people. What is probably necessary is some sort of moderation system, preferably one that involves the users, not the employees of Friendster. The backlash against the clamp-down on the fakesters is just an example of the fact that the users of Friendster don't want babysitters to sanitize the experience for our protection - for that, there's AOL.