Every Piece of Information Is a Latent Community
December 2, 2008

Clay Shirky’s keynote talk to open the 2008 Online Information Conference

“group action just got easier” = 5-word summary of his book Here Comes Everybody

the ways the media environment is being transformed now that consumers are first-class participants
the overlap of all of the patterns in one environment is the big transition we’re all living through and trying to figure out

showed a picture of a truck in a parking lot at sunset on Flickr - HDR photography (technique, not just software)
don’t need to see what’s going on in the comments to understand what’s going on there
people start inserting photographs into the comments, which turn to a technical discussion
a user group is assembled on the fly
used to be gather then share - used to have to identify the people who would be interested first and then organize/share
Flickr reversed the pattern - share and then gather
they didn’t identify themselves before they saw this page
Flickr had the infrastructure to let these people create a community on the fly
once the users created this, it wasn’t evanescent anymore - it was permanent now

shows that every URL is a latent community - potential value that people looking at it might find value in it
not all will see community grow, but the potential is there
can have many more communities of practice at much lower cost because the old distinction between conversation and publication is no longer true
why pick? Flickr gets more value out of not having to decide in advance what a piece of information might be used for
even on the Flickr picture, other conversations can take place in parallel

Flickr gives users the tools to add value

there are large patterns we see (not every service on the internet has these, but some large ones do)
- share
- collaboration
- collective action
in this order, because how much does the individual have to give up to get value?
takes more effort the higher you go on the ladder

showed Bronze Beta - the Buffy the Vampire Slayer fan club site
back when WB sold the rights to Buffy to UPN, UPN didn’t want the community group online, so they shut down the server (UPN: we don’t want it because we’re in the television business)
the users, however, wanted the community to continue, so they raised money and commissioned a new service to move to
they explicitly decided they didn’t want any “features” - no ratings, rankings, etc.
they just wanted to type in text, and now it’s just a giant scroll of conversation
the community is still going

these new social technologies are the first time where later generations of technology have fewer features than older versions
the simplicity in the tools has to do with a mindshift of the computer as a box to a door
for individually-oriented software, a long list of features is good (Photoshop, Word, etc.)
but when we want to collaborate, fewer features is better; we need the same mental model of what’s going on
the complexity is in the user, not the software
in Bronze Beta, the complexity is in the very long list of rules created by the users (”no colored fonts”)

showed the Wikipedia entry for Doctor Who - it’s been edited almost 9,000 times by more than 3,000 people
the breadth and depth of participation is quite extraordinary

“hive mind” - people that use this term almost always don’t understand what’s happening
these folks aren’t part of a community in any sense because most have only edited it once or twice
someone, though, has edited it thousands of times; every article he’s touched on Wikipedia is about Doctor Who
there is no coherent average behavior, although the commonest behavior is one edit, one user
we’re used to counting noses - how many people watched my TV show, read my book, etc.
but here, there is no one common user behavior; instead, there’s this tiny group of fantastically engaged users
imagine going to your boss and trying to convince them to plan this
it’s not everybody pitching in like a barn-raising; it’s not collaboration
it’s like a small, self-appointed editorial board

collaboration involves real synchronization
it’s not just you share and I share
collective action is the most difficult pattern to get going because the whole group has to commit to it and either stand or fall together

two examples - HSBC

they recruited college students with penalty-free checking accounts
proved to be popular, but then they changed their minds and added a penalty
gave users 30 days to get their money out
thought they had the information and coordination advantage
in the summer, the students should have been outclassed by HSBC’s tools
but they didn’t count on Facebook
a user starts a page, which goes viral
for the first time, college students are dispersed but active
they started sharing documentation - good banks to move to and how
once one person solved the problem, the information was available to everyone
goodbye to HSBC’s information advantage
then they organized a real-world protest, but it never happened because by then HSBC had caved in
HSBC backed down because the students were upset AND coordinated

“thinking is for doing”
there is an analagous transformation that publishing for acting
the newspaper could only report HSBC had changed the deal, while Facebook could actually encourage users to do something
publishing and action is no longer a choice - can do both
now have a response without managerial control

example two - flash mobs
they were promoted in “emails by bill”
wanted to prove that hipsters would do anything you told them to
hits belarus - eating ice cream in Minsk Square
the police showed up - the group became a problem (not the group eating ice cream)
it had been made illegal to act in concert - to be a group
when they entered the square, they weren’t a group
the livejournal page led to action - it’s a full cycle; they didn’t just bring their ice cream - they also brought their cameras because they wanted to document the state’s response

in less than 3 years, flash mobs went from being something to mock a certain class to political protest
we tend to underestimate the potential of these tools because they tend to look frivolous
we don’t understand their potential
anything that allows group formation is political
so much of the meaning of the tool is in what the user does with it once it becomes social

what is all of this doing to the media landscape as a whole?
we’re living in the middle of the largest increase in the social expression of the human race

1 - printing press/movable type
2 - point-to-point communications (telegraph, telephone)
3 - capturing sound and video
4 - broadcasting spectrum (radio, television)

curious asymmetry to them - the ones that are good at creating conversations are not good at creating groups and vice versa
there was no medium for creating two-way conversation among groups (many-to-many) until now
there is no longer a distinction between consumer and producer
giving someone the ability to receive email means they can send email
the audience grows and becomes varied
the 5GB generated this year will be at the edges

the internet is also the mode of carriage for all previous media as it’s digitized
it’s also adding social dimensions to all existing media

to produce something for a lot of people to watch, read, etc., I have to take on a big burden for production costs
if I’m wrong, I lose a lot of money
in an era of gutenberg economics, I decide which books are good and I publish them
all following media have had the same economics problem
filter and then publish becomes the model - see what’s good and then publish it

now, anyone can publish to anyone with a marginal cost of zero
it’s the first medium we’ve had that works with post-gutenberg economics
anyone can say anything to anybody and they frequently do
it’s too much content to filter in advance, and there’s no economic reason to do so

the question for a 15-year old today is not “why publish” but “why not publish?”
many of the huge businesses built on the back of the internet have at the core of their business model a post-publication filter
get to the good stuff after the fact, not before

the users are now well and truly engaged in the publishing environment

the user as publisher model:
1- Gnarly Kitty
a fashion-obsessed Thai student who posted about a fishing game
why would anyone publish that?
because she’s not talking to us - she’s talking to her friends
we’re not used to seeing things that are public but not in the public
then a coup happens in Thailand, and the government tells the media not to report about it
but Gnarly Kitty publishes the first picture of tanks in front of the parliament house and she is now the go-to source
people are now flooding in and she becomes a global resource
then she posts about a phone she’d like to own
the users get upset and want more about the coup
she responds with a post that it’s *her* blog and it’s about her life

zuckerman: journalism has gone from being a profession to being an activity
she committed acts of journalism; she just did it while she was a concerned citizen
not connected to self-definition
this model is new
she doesn’t need the money to be a global publisher
she gets thousands of new readers and she tells them if you don’t like her content, then leave

2 - Howard Forums
early blog about cell phones
can”t answer people’s questions about their phones, so he says hey, you all talk to each other and he puts up a forum
is up to a billion pages this year because the expert users are solving problems for the new users
tech support reps from phone companies will refer callers to the Forums
they have access to “reality,” which the engineers don’t
the kinds of questions that can only be answered when A has part of the answer and B has the other part and they collaborate
users creating detailed technical documentation
it’s not all tech all the time, because users have gotten to know one another and they hang out here together (they post pictures of their pets)
as a publisher, it’s easy to see that you’d get rid of the pictures of cats
but that misunderstands what is going on here
that both of these things are coming from the same web
they’re not doing one in spite of the other, but rather because of it
it’s the fact that the users care about each other is what gets them to do all of this
communities have to be for the members
the satisfaction comes from membership and recognition from the communnity
hosting that isn’t amenable to crowdsourcing solutions
communities need to get to know each other and share all kinds of things in order to do the technical documentation

3- showed a still shot from Joss Whedon’s new show, Dollhouse
fan experience is that his shows get canceled, so they’ve already created a site to save it from cancellation before it even airs
in the past, they’ve organized protests
they don’t trust the marketing department to explain to people why they should watch it, so they do this themselves
there is no aspect of the information industry that users aren’t crawling into, including the marketing department
users don’t always do this well
the pattern is usually extract the signal after the fact
they do always do it differently, though
grappling with that difference is the big question we have to deal with now

one of the big changes is that anybody in any part of the information business is now part of the entire information business
no longer i work in television and you work movies - it doesn’t matter anymore
no longer that we produce the content and then the users go off and talk about it somewhere else
creating community and arranging action are now part of production
not every organization should get into every part of the business, but publishers can now be conveners of community
can allow amateurs in to extract value - that’s what we’re grappling with
it’s not a move from A to B but from one to many
the landscape itself is expanding

when the printing press came out, it wasn’t that people looked at it and said, oh now we need a printing industry and this is what it will look like
little things turn out to be big deals
making books smaller meant more people could carry them (creation of octavo size)
if it’s hard for a thief to get a book out the door, that’s a feature
that little intuition sparked a revolution

everybody is everywhere and all the walls have fallen
everybody can see each part of the business; it’s all horizen and no barriers
what’s the next good thing to do?
the answer is most certainly to explore
experimenting our way into the future is what will show us what works
there is no roadmap for the period we are entering

q: what is the role of the professional librarian
a: liz lawley says libraries are “happiness engines;” the whole of the world that deals with traditional publishing is now dealing with the split between lovers of the page and lovers of the book; it’s easy to see the role of librarians as hosts of books, but if you see sociable libraries as happiness engines, then the question becomes what set of things done in libraries now would increase the happiness; one of the obvious answers is “collaborative filtering” - helping the user find the next thing to read, watch, etc.; libraries have typically serviced users one-to-one, but there are groups of people coming together and talking with each other in the library; ideas make people happy, so what resources do we have to extend that; one of the big resources we have is that we have “convening power” - it’s unmatched in civil society; the cross-section that goes into a library is quite extraordinary; it doesn’t have to be one-to-one, and there is a great deal of potential in experimenting with many-to-many; even in the corporate world, libraries can join up people who should be talking with each other; IBM example - “DogEar” plus a one-way mirror; allowed researchers to tag URLs, although they’re not sharing the tags back to the world; two geographically-dispersed research groups there discovered each other because they were tagging the same resources, clearly with the same ideas; they actually called each other and then pooled their efforts; this would never have happened from the top-down; “research is a famously upside-down problem” so there’s no way one person at the top could have said these two groups in two different countries will work together; when the users can see what each other think (don’t apply the ontology in advance), people with similar world-views can be connected; connecting users because they’re looking at the same information

q: if we spent our lives organizing information as a community, how do we tackle all of the new information being created?
a: you can’t; you only have 2 chances to actively organize things - moment of creation and moment of use; at creation, can try to add metadata, but at use stage, you can involve the user and have them modify or verify the metadata; the problem becomes a little bit of effort gives you a high degree of leverage, so have to find the right point where this happens; there’s no way to apply the metaphor of the shelf to cyberspace; they have to do with automatic extraction, inviting users to upgrade metadata at the point of use

q: what does this tell us about human nature that we might apply to things we do?
a: that is THE question, in part because it’s the one we need to answer but can’t; used to think that the world was changing because technology was changing, but now thinks we’re just not used to explaining human behavior without being paid or other extrinsic motivation; we used to think the market was the public sphere and the household was the private one, but that’s changing; Wikipedia makes no sense at all; what critics have missed is that human nature contains an enormous amount of Gnarly Kitty motivation;public and private sphere are existing side by side, can’t be explained purely by the market

q: the idea of expertise as opposed to popularity
a: if your skull is going to be cut open, you want it to be done by a trained professional; the reverse is that you don’t need to buy music only in the presence of a record store professional;
the closer things to come to life and death and one-off decisions with no reversability, the more we want expertise; the places where there is an obvious right answer that is independent from the social view; changes here are coming about in the end of the spectrum where what people believe changes what is true; are SUVs a truck or a car? that decision was socialized, which got us to a better answer than letting Washington decide; there’s no general “get ouf jail free” card for experts; very often, the really interesting hybrids are where professionals and amateurs come together; in most but not all cases in the information industry, it’s headed to hybridization because it’s not the critical one-off decision; how many different strategies can we apply to see where the cost versus value curve is

q: should we be worried about efficiency? should we be worried about experts? one of the problems of community is that there are maturity issues that affect newbies (keep learning or does everyone become an “expert”)
a: the social origin of good ideas; putting experts and amateurs together improves both groups because when the expert has to teach, he learns; it’s the conversation between the two turns out to be more powerful than pure amateur aggregation or pure expert knowledge; these systems work not because they’re efficient because they’re effective after many fruitless tries at low cost; resources don’t get tied up in the failures because it’s easier to identify them; we’ve all been in that meeting where we realize we’ve expended more energy talking about the idea than we would have if we’d just implemented it; most Flickr pictures don’t have comments but it doesn’t cost Flickr anything; that’s why these new systems look so strange to us

The transcription factor c-Myc is essential for cellular proliferation and is one of the most frequently activated oncogenes, but the...