stimulant

Why are my friends and acquaintances significant to me? Largely (and couched with deliberate callousness), I care about them because of what they do for me. They provide value. Whether this is because I inevitably have fun around them; because they know how to do the things I want to; or because they have access to ideas or people or equipment that I need; the underlying theme is the same: my friendships are determined by my interests (in the game theoretic sense) and how my friends (and those resources to which they have access) help me pursue my goals.

Despite the inflammatory wording, this is true of everyone. Maybe some people would suggest that friendship is unconditional, but nobody would recommend staying in a net negative relationship.¹ Everyone uses their friends for something:² sex, fun, emotional support, money, homework help, etc.

My social network is just the collection of people I have found useful enough to connect to.

social networking still hasn’t left the gate

So, this is what my real social network does for me. What about my online social network? Well, it can tell me who I know in the real world, where they live, what movies they like, what they’re interested in, whether they’re up for casual sex, … in short, it doesn’t tell me much. Sites like facebook provide for pretty terrible network searching and browsing if you don’t know who you’re looking for, and if you’re not searching within a small, particular, explicitly defined³ group. Honestly, I don’t care that much about what movies my friends watch. It’s good to have access to this information, and it’s certainly valuable. But, there could be much more potential.

People commonly cite the idea that “social networking is a feature, not a product” as one of the hallmarks of “Web 2.0.” But in this regard, social networking hasn’t even left the gate, having failed to realize that there are real, significant reasons we want to connect to other people that can and will exist outside of social network software. Furthermore, contexts exist other than those manufactured by sites like twitter or iLike or Last.fm; in fact, those are the least important reasons I care about knowing somebody. Right now, social network software focuses on who your friends are, rather than why you’re friends with them, despite the fact that it is the latter information that will allow social networks to enable really significant collaboration. This is why calls for portability, openness, and interoperability make so much sense. Despite these possibilities, we’re still connecting people together for, in many cases, no particularly good reason. The social networking “workflow” ends with your connecting to or finding the other person (or maybe peters off with your ability to message them or poke them or dedicate a song to them). But, it can and should begin when you “friend” a person: finding someone is only the seed, the first step of community building. Once we acknowledge this, we can envision some pretty neat possibilities.

identity

Consider the thorny problem of identity and anonymity online: right now, we manage our identities — that is, we disambiguate and distinguish ourselves — by creating and maintaining the privacy of secret information (most frequently, usernames and passwords). But, in the real world, this is not at all how we operate. More fundamentally, secretive, unique data are not what actually define our identity: they’re artifacts of our interactions with massive, top-down organizations like schools and government. Our real world identity is defined by whom we know, what we do, what we read, what we like, what we dislike, etc. Our identities arise from how we apportion our attention and to what resources — people, things, knowledge — we have (and conversely, want) access.

As more and more of our life goes online, we are looking for increasingly dynamic online environments and platforms that are responsive not only to our needs and our explicit requests, but our identity. Historically, browsing the internet has been a passive process wherein the user is a spectator moving among static pages. We’re moving toward a participatory society⁴ where our interactions are increasingly active: they no longer comprise simply receiving information. This active component of our online experience goes beyond content creation and extends to a fuller recognition of our identity. Commensurately, we’ve been giving up more and more of our anonymity. This is a natural consequence of moving from consumers of online information to producers of online content and actors in the online community. What does this mean for the management of our identity online?

Currently, identity is managed pretty shallowly. For most rich internet applications (RIAs) and communities, identity is managed with a user account. Depending on the information you give the application and the preferences you set, your user experience changes. Unfortunately, the process of giving applications information is still relatively labor-intensive. Although people have recognized⁵ that if you can convince users to give you loads of ostensibly useless, raw data (e.g. browsing history or bookmarked sites or book preferences) about their lives, you can almost always extract and monetize a useful application from it, we still conflate the application and the collection of raw data. This practice naturally fractures users’ online experiences, requiring them to set up redundant user profiles, and in turn handicaps developers who always want more data.

We are still stuck on the idea that we need a unique identity in all places. For example, when you log into Amazon, you need to see your recommendations. What if your results overlap 99% with twenty or thirty other people? You could gain increased privacy without trading functionality if you realize that your identity is not an all-or-nothing piece of information: you can reveal your book preferences without revealing your name and credit card number. In the real world, there are a variety of ways to control how much of your identity you reveal. Social networking could provide a robust sense of identity if we expand our sense of what a node can be on a social network.

What if it could be a movie, a place, a body of knowledge, or a person? Then, the information in your social network would become a deep but very flexible representation of the what that you find pertinent about yourself. Precisely the facets of your identity you’re interested in sharing with developers would be made available, and their availability would itself be flexible. Essentially, by allowing for the addition of markup or meta-information to the edges on your social network (i.e. your connections to other people), one could provide for levels of privacy, or for easy access to subsets of anonymous but useful information (e.g. what restaurants you like) which could then be made available without revealing the other elements of your identity that uniquely define you. Rather than solving the privacy problem with more security, we could make an end-run around it by more finely controlling what information we reveal. Note that this is a fundamentally flexible and dynamic decision: rather than a static user profile requiring updating, a user can simply choose to reveal more information. No one need engage in the painfully redundant process of writing up user profiles and accounts.

For this dream to come true, companies and services (both online and offline) need to collaborate to create an open, independent network. Frankly, I don’t know if facebook is up for it. I’m a little flabbergasted by the dearth of collaboration between companies, particularly online! Given how much of a service’s ability to provide deep user experiences depends on the information available about its users, openness is clearly in everyone’s best interest. The fact that an API and a set of closed data is the best we can do is concerning. Luckily, people are asking for more.

network of trust

Calling for such openness naturally raises questions of security. If we broaden how we think of a social network just a little bit, we can take care of that, too.

In May of 2007, Linus Torvalds gave a Google Tech Talk about source control management (SCM) and in particular, his SCM solution, git. Version control is a really, really powerful idea that confronts a lot of the problems central to collaboration. Those problems remain, for the most part, unsolved; despite the fact that there are many domains this solution could revolutionize — like Congress — it’s not quite what I want to focus on, today. I do want to take a look at an excerpt from Linus’s talk (emphasis mine):

In practice you will never see, oh, there will be a thousand or maybe twenty thousand different branches, but in practice you won’t ever see them because you won’t care. You will see like a few main branches, maybe you’ll see only one. in the case of the kernel, a lot of people, they only ever look at my branch. so even though there are lot of branches you can ignore them. What happens is that the way merging is done is the way real security is done. By a network of trust. If you have ever done any security work, and it did not involve the concept of network of trust, then it wasn’t a security work. I don’t know what you were doing but trust me, it’s the only way you can do security, and it’s the only way you can do development. The way I work, I don’t trust everybody. in fact I am a very cynical and untrusting person. I think most of you are completely incompetent. The whole point of being distributed is I do not have to trust you, I do not have to give you commit access. But I know that among the multitude of average people, there are some people that just stand out and I trust, because I’ve been working with them. I only need to trust 5, 10, 15 people, if I have a network of trust that covers those 5, 10, 15 people that are outstanding, and I know they are outstanding, I can pull from them. I do not have to spend a lot of brainpower on the question. […] And the nice thing about trust is that it does network. That’s where the network of trust comes in. I only need to trust a few people that much. They have other people, they have determined, hey, that guy is actually smarter than I am, that’s actually a really good measure of who you should pull from. If you have determined that somebody else is smarter than you, go for it. You can’t lose. Even if it turns out that you pulled crap and somebody else starts complaining, you know who you pulled from and you can just point to that other person and say “hey, I just pulled, go to him, he knows what he is doing”.

Linus is only talking about development, here. But, SCM is a powerful idea⁶ , and in particular, the idea of a “network of trust” seems underimplemented. Social networks provide an extremely natural way to generalize the idea of a network of trust and make the feature modular and portable.

Currently, social networks are very simple graphs whose nodes are people and whose edges represent acquaintance of some sort. Our real world connections between people are much subtler and much more multi-faceted than “acquaintance of some sort.” Unwittingly, as far as I can tell, people have been slowly straining to add more information to these edges: e.g. facebook allows you to note how you met someone, while the Top Friends and Enemybook/Snubster applications clumsily let you note what kind of relationship you have to a person. But this is really a joke compared to what we should be expecting of social network software. What if we could define attributes by which we classify our connections to people, defining subgraphs of our social network to which we can attribute privacy settings, levels of trust, etc. What if rather than being added to the developers’ permissions list for a version control application, you simply joined a developer’s group online? Rather than leaving the definition of groups up to the service, what if you could decide what groupings were important, and make these distinctions available on-the-fly to other applications and services? Again, the idea is that identity exists t many more scales than we acknowledge currently, and we not only need to recognize those scales, but give users finer control over what scales — i.e. what facets of their identity — they reveal.

needs

• We need to get over the newness of the idea of an online social network and reappropriate the knowledge and intuition we’ve already created for designing and capitalizing upon social networks.
• We need to broaden our definition of what constitutes a node on a social network so that we can start adding extrapersonal significance to the documentation of our relationships.
• We need to understand the long term benefits of openness: rather than trying to capitalize on incremental innovations in the short term, we need to work together to assemble a flexible infrastructure for the long term.
• We need to start identifying the themes behind people’s problems and desires (e.g. resource matching!) and make tools to address these; domain specific problem solving can lead to an unjustified lack of interoperability.

I’m excited!

1. In fact, it’s hard for me to reconcile the existence of an affective metric with the idea of unconditional emotions: the metric needs to be fully irrelevant! [↩]
2. I understand that I’m bastardizing the meaning of “exploit” and the connotation of “use,” but please bear with me. [↩]
3. By which I mean that users have set out an explicit group (e.g. a class) as opposed to an implicit group formed by facebook on the fly, given desired parameters [↩]
4. Some would say that we’ve arrived; however, I think we’ve a long way to go [↩]
5. I know I’m paraphrasing from a source elsewhere, but I can’t seem to find it again: any hints would be appreciated [↩]
6. If you skipped the link the first time, go now and read about Congress’ need for version control. [↩]