2004-02-11 Revenge of the User: Lessons from Creator/User Battles http://conferences.oreillynet.com/cs/et2004/view/e_sess/4948 (danah boyd) her blog: http://www.zephoria.org/thoughts/ social hacks [it's adaptive design.], what users do. how to people get jobs? how do they make friends? how do they connect to poeple they know? technologists saw this and tried to systematise this. eg friendster. didn't really work. people have private ways of working with their social networks. but technologists took this and made something public, very different. so the users took this and did something different. technology shifts the social architectures that we use, it changes what we actually do. code is law, code is architecture. [bingo] so, example: real life spammers. we structure our lives to route around these people. we don't go to those bits of the cities to not get pushed xxx videos at. but friendster is a public forum, the same for everything. [no distance. no ways of attenuating differences/difficulties/mistakes.] our model is: everyone can be connected by 6 degrees. but we're not *actually* dealing with top-down networks. social networks you route the parcel by giving it to who you *think* is the best next connection. we have no idea of the *actual* number of degrees that separate us. robin dunbar and 150. [damn. i'm talking about that. got a whole slide on it.] people groom and support their social network with gossip. but this 150 number isn't the number of people you ever encounter. it's a dodgy theory. another theory that gets used: strong/weak ties. jobs usually come from weak ties. so we assume that a friend of a friend is a weak tie -- but that's not the case at all (necessarily). social network theory also talks about bridges etc. knowing about a job from a weak tie is a power thing, that's what bridges do. the closer you are, the more you have in common [the closer you are, the more likely you are to have the same information IO.] friend of a friend needs to take into account context and how much you have in common: if the friend is a hairdresser, probably not good. but if you're both in the tech industry, foaf could be good. so, apply this to the social networking sites: 1. friends on these sites are no close ties. so friends of friends is really weird 2. favours are odd. people tell you about jobs because they like to be bridges. asking favours is very different from offering them. because social networks are explicit online, that puts a lot of pressure on people to do introductions [there's no plausible excuse to say 'i can't do you and introduction' if there's a context problem: work/play] 3. they aren't your friends, they're your friendsters. but you have no idea how important someone is to someone. there are significant problems articulating who you know -- try saying "no, you're not my friend" to your boss you want to get on with. relationships aren't two-way. you can't ask a shrink the same questions they ask you. these networks are socially awkward, they don't feel real. context clash: a schoolteacher who is friends with both students and RL friends. everyone has a different metric for how to solve this. linking to only their really close friends. how you sort out their profile. (people's profiles are written the first time they came on. so it looks like friendster is all burning man freaks.) fakesters was one of the things that came out of this to help. a whole bunch of professors dropped out of friendster and made a fake account to hide from their students. [so the distance that knowledge transfers over (and evaporates) is done in the real world again, by moving the knowledge of that identity being that RL person leaves the digital world and goes back into normal social networks.] what these sites are trying to do is configure the user. the technologist creates an object that has an explicit boundary. the technologist force people to conform to that. [yes yes! configuring the user is fordism. is another pointer to the fact that computing is another unfolding of the industrial revolution mindset.] [i wonder whether people walk to the left of the right of the room depending on which side of their brain is strongest? have we self-divided? the right side of the room doesn't look at cool as this side.] final lesson is, these things are different architectures. they're not real social networks, but that's cool too (except that the business models don't like that.) what we can see/questions: 1. how do we create a naunced way to do social context without creating uncomfortable situations. 2. how to let people show face? you talk to a 3 year old and people you work with differently? you need to be able to politely lie. livejournal lets your set who can see what posts. 3. how can you stop people from being the subject of hate attacks? [it's the social distance thing again] 4. how can we create a social architecture that allows regulation by social norms. allow social to regulate social. and that last one is the big thing. why are users misbehaving, and how to learn from that? q about how friendster changes existing RL communities. interesting answer: people used friendster as a spam-free way of sending email (a 1/3 of people use hotmail as an email client). you can click on someone's face and mail them. [completely. why does my mobile have number buttons on it? i barely ever talk to people i don't know.]