BlogTalk Reloaded 1.1
I’m going to attempt to blog as much of the conference as I can. This is the usual caveat about the fact that what follows is my notes from these talks; any flaws in my representations of papers or conversations are mine, and not those of the presenters.
Opening keynote: danah boyd, “The Significance of Social Software”
what is social software? in 2002, shirky claims term to apply to “all uses of software that support interacting groups, even if the interaction is offline” (Allan)
applications of term focused strictly on new, ignoring previous forms of online interaction (muds, moos, etc.); also ignored corporate groupware uses
shirky: social software as “anything worth spamming”; tom coates: “software that supports, extends…”
examples: blogs, social network sites, media sharing systems, wikis, tagging, mashups, etc
people got annoyed: ignored history and important older forms
danah restricting her use of term to post-tech bust uses (“web 2.0”)
three changes created by social software: design, spread of participation, social behavior
argues friendster killed the previous uses of terms like “beta” and “version 2.1” — moved away from engineering model of software releases to a definition of beta that meant “not yet profitable”
traditional design process: slow, bureaucratic, painful; many entrepreneurs and developers said they didn’t want to work that way
folks behind myspace decided to do friendster better; hacked together a piece of software based on coldfusion, invited friends, people who got kicked off friendster, etc; no spec, no legal, no marketing — just shipped (in 2 months!)
what happened next? shipped, and then asked “let’s see how people use it”; asked users what they wanted; launched feature after feature based on what people want — all hacks, not traditionally developed; no quality assurance team
set of values of social software world:
— hack it up, get it out there
— learn from your users; evolve the system with them
— make your presence known, invite feedback
— monetization? add a few ads here and there
pros and cons to this approach
— horribly unstable code
— but gets outside the lab-driven context of human-computer interaction; focuses instead on human-human interaction
what happens when crowds come is very different from what individuals do; anything that can be fucked with, will
values are built into software
example: flickr — developers introduced themselves to all early users, said hello, asked why they were there and what they wanted; created context of community
such values are often tech-centered, at least at first (see delicious, digg)
where things go wrong (?): orkut (part of google; known as brazilian site?)
originally invite-only, mostly tech folks; quickly went to bloggers, but by that point, bloggers were sick of social network sites, so they paid little attention; some bay area brazilians were invited, they invited more brazilians; some kind of flag-based competition that drove more brazilians to join — google was clueless
later tried to spread to india — has now replicated caste structure — google still clueless
ways that cultures set norms for behavior in spaces
impossibility of speaking in multiple registers at once — example of stokely carmichael
terror in us about myspace; teenagers behaving like teenagers, not attempting to communicate with adults around them, alarms adults
social software shifts internet use from interest-driven (usenet model) to friend/community-driven — but such a system begins to fall apart when the scale gets too large
example: facebook — moving outside the college/university context caused protest and anxiety
how far can things scale?
interestingly, blogs are the only form of social software that has scaled well, in part because each blog is individual, and networks that form among bloggers are separately developed
— designers — what are costs of chaotic design processes, and how can the processes be used most effectively?
— researchers — what are the implications for people’s daily lives (?)
— busness folks — are there ways to rethink scaling process without killing communities?
tendency to celebrate new and forget old has costs — things social software folks can learn from older forms like MUDs and MOOs
No mentions yet.