Social informatics and social intelligence

[Hello Rutgers Online people! Leave a comment!]

Ricky delicious’d me “Social Computing: From Social Informatics to Social Intelligence” by Wang et al t’other day.

After introducing the basic idea that “social computing, a new paradigm of computing and technology development, has become a central theme across a number of information and communication technology (ICT) fields” the authors go on to give a few definitions of “social computing”, all of which beg the question to varying degrees.

Calling something social software (which is the term of art I am more familiar with) implies that there is also just software. Software which is explicitly or implicitly not social. Julian Bleeker thought about this very topic and said, basically, that all software is social. Constructing software is a social activity among hackers. Using software, even something “boring” like Excel, is a social activity. But, at the same time it is easy to see that there is a new category of “social software”. So, social software, in Bleeker’s term is software that explicitly has an emphasis on creating, viewing and using “network-based articulation[s] of social formations”. This is an important distinction.

Wang et al’s definition is the closest to the idea that social software is something that makes social formations clear:

[Social computing and social software is] computational facilitation of social studies and human social dynamics as well as the design and use of ICT technologies that consider social context.

Most current “social software” is of the sort that Bleeker describes: something that makes social relations clear, or at least allows people to work with social relations in order to facilitate something. At it’s most simple, this takes the form of the “friend list” on sites like Flickr, Twitter and MySpace. People who you “friend” (and graduations therein: Family and Contacts in Flickr having higher and lower weight respectively) are allowed slightly more information about you than people outside the social circle. They can do more for you and more to you. However, Wang et al seem to be going down a different path than most people interested in “social software”. They’re talking about:

the representation of social information and social knowledge. Other important issues are the modeling of social behavior at both the individual and collective levels and analysis and prediction techniques for social systems and software

Whoa! This requires some unpacking.

The representation of social information and social knowledge. How far do you want to go? I have the feeling Wang et al mean something far beyond social-network diagrams. They say “Social information describes societies’ features, such as social relations, institutional structure, roles, power, influence, and control.” Well, yes, and more besides. I’m sure they’re aware that they’ve just smushed all of the social sciences into one sentence but I’m not sure what there is to gain from representing social information and social knowledge except for being able to model it.

Modeling of social behavior at both the individual and collective levels. Modelling anything is fraught with difficultly because the model is not the thing, it’s an abstraction (which Wang et al state). They’re going to model social behavior with agents. Autonomous agents (which is clearly a term of art) to be specific (is there another kind?). But, why do you need to model something? To analyse and predict.

Analysis and prediction techniques for social systems and software. Wang et al say that you can use statistical techniques to analyse and predict all sorts of things. True. And in the physical sciences you can do this with great accuracy. But when you start trying to analyse and predict things like “the costs and benefits associated with various strategies, policies, and decision-making methods” you run into economics and other social sciences. Economists, using the same (statistical!) modelling techniques, rarely agree on anything, hence the joke “if you laid all the economists in the world end-to-end, they’d still never reach an agreement”.

Wang et al also say “Social computing enables building social systems and software and allows for embedding actionable social knowledge in applications rather than merely describing social information. ” (my bold) I’ve read the article twice and I still don’t know what actionable social knowledge is or why I’d want it in my software.

Wang et al conclude with “From both theoretical and technological perspectives, social computing technologies will move beyond social information processing toward emphasizing social intelligence” which really pushes my buttons. I am very much on the soft-AI side of the fence so whenever someone starts talking about technology having some sort of “intelligence” I come over all Cayce Pollard.

In fact, I’d argue that I very much do not want “intelligence” in any software I’m using or even “actionable social knowledge”. And that argument will be in my next post.