Clay Shirky closes with some observations. First, too many social spaces online are either too big or too granular in the level of discussion and interactions they promote. What he and his graduate students are working on is to try and describe common patterns in asynchronous social spaces (like Slashdot).
Here’s how he represents Slashdot’s mechanism: get dedicated users to protect readers from writers via various strategies (moving comments to a separate page, mark off writers and readers as separate, let users rate posts, and “defensive defaults”.) He observes that in these spaces talking about “users” improperly confuses very different entities within the site. So then how do “guard the guardians”? Treat users and members differently, measure good behavior, enlist committed members, and keep judges from posting. Clay argues that this helps to keep techies and non-techies from strong differentiation. Clay also claims that what this does is work towards a goal of annotation and against conversation.
Second example: Bronze Beta, a Buffy the Vampire Slayer fan forum. The users took this one over and said to the designers they hired, “1. Don’t have any features. 2. Make comments central. 3. Make login optional”. So they don’t even have threading or topic lines; the whole site rolls up onto one page, and that is the chronological record of a single continuous conversation.
Third example: Writely. Writely “treats the group as the entire user”. One login for the group, by invitation only, time limited work.
The project Clay is describing is being discussed at the Pattern Language Wiki.
Danah Boyd’s talk: when does the social trump the technological, when does the technological trump the social?
Argues that you don’t really want to attract everyone indiscriminately to a new piece of social software. What you want is for a specific set of people who find each other in the space, and become devoted to using it. Mentions the inflow of people to the early organic growth of Friendster as polluting its uses for those who had adopted it early. [Reminds me of when AOL users flooded into Usenet.]
The problem of organic growth is “echo chambers”, what she calls “cluster effects”. You get all the same people talking to each other. [In my view, this means they have a low ceiling on what they can accomplish or do in the longer run].
Discussion of MySpace in this context. Argues that distinct clusters enter together, originally out of connection to music, but that once they’re in, they can rub up against other clusters who also joined as a group.
Now we’re collecting cases of social computing where there was a huge disaster of some kind. Examples: 1) Sims Online 2) The LA Times Wiki-editorial 3) An early design for instant messaging at Microsoft that was a full-screen interface. 4) Cobot 5) Howard mentions a philanthropy example I haven’t heard of 6) Star Wars Galaxies. After this we started moving into much higher-level abstractions about the impact of social computing.
Final comments from Elizabeth Churchill and Wendy Kellogg. Summary of themes over the last three years: social networks, blogs, tagging, teenagers, cross-cultural problems, politics and collective action, virtual worlds, games, the definition of ‘community’.