Full Interview: Clay Shirky

Have you ever played around with a gadget or application, only to discover it’s absolutely perfect for something different from its original design? This kind of inventiveness, or playfulness, happens all the time in our digital environment, but it signals a major shift in the relationship between the inventor or designer and the user.

Nora interviewed Clay Shirky about just that earlier this week. Clay is a big thinker on internet and culture, and he has a lot to say about how users shape the tools they use and how designers should respond to this new “interaction loop.”

A shorter version of that interview will air on Spark 93, but you can listen to the whole thing now, or download the MP3.

Play audio:

If you like hearing these extended interviews, why not subscribe to Spark’s enhanced podcast? You’ll get regular weekly episodes, plus additional blog-only content like this. [Subscribe via RSS] or [Subscribe with iTunes]

[Original image by Joi Ito]

Full transcript after the jump

Full Transcript

Nora Young: Hi, I’m Nora Young, host of Spark on CBC Radio. This is my full, uncut interview with Clay Shirky. Clay is the author of “Here Comes Everybody,” a book about mass collaboration, and he’s also a professor at NYU’s Interactive Telecommunications program. An edited version of this interview will air on an upcoming episode of Spark, but for now I hope you enjoy this full conversation. Hi, Clay.

Clay Shirky: Hi, Nora!

Nora: So, your talk at the TED conference earlier this year, you started with this example of how people use cell phone technology in the US and Nigerian elections. And in the US, people used phones to video polling stations, making sure there was no voter suppression going on; and in Nigeria, people used cell phones to send text messages during the election and sent that online. So, what does that example show us about how innovation happens now?

Clay: I think the important thing about that example is that the social pattern of involving people in monitoring elections is somewhat independent of the tools. It’s not completely independent, you have to have some tool to do this kind of shared monitoring, but it’s also not completely dependent, you don’t need the fanciest cell phone. And a lot of what we’re learning about design of these social tools is, once you understand a particular pattern you can start to replicate it in all sorts of places. Even though the original conditions where, again with election monitoring, even though the original conditions of election monitoring aren’t the same everywhere in the world you can recreate some form of election monitoring once you understand the social tools you need to do it..

Nora: OK, so when it comes to this kind of spirit of innovation spreading, what are some of the conditions that you need to foster that kind of spreading of this social innovation?

Clay: You need a very low cost of experimentation, right? If things are expensive to try people will hold back from trying them and they’ll spend all their time trying not to fail. If the cost of experimentation falls though, and I mean falls precipitously, then people will spend a lot of time experimenting, and instead of not failing, the goal becomes to fail informatively to learn something from the things you tried. So, Ward Cunningham, who created the Wiki form, launches it in 1996, and it’s mainly sort of groups of programmers using it, but anybody who looked at the Wiki pattern could essentially write their own Wiki because it was the conceptual shift that Ward had done. The software wasn’t terribly complicated, it was really the conceptual model of how could people trust each other in a collaborative group.

And so, after five years of that, of famously, you know, Jimmy Wales and Larry Sanger adopt it for what became Wikipedia. But, by that point, both the way the tool worked and the social pattern around it were well enough understood. You couldn’t I think have gone straight from Ward’s original Wiki to Wikipedia. There was that long period of low?cost experimentation.

So, I think it’s at the technical level, low?cost of experimentation, at the social level, a high willingness to share results. And those two things create a kind of feedback loop that’s created a lot of this surprising pricing progress we’ve seen.

Nora: And so is that why the Web in particular is a good format for that kind of innovation, because the barriers to experimentation are so low?

Clay: So, there is a real chicken and egg question. I mean there is a doctoral thesis in that question, and it’s not the one I’m going to write. There were two different things about the Web. One, the barrier to entry is very low, exactly as you say. But, two, the culture of the Web is absolutely predicated from the earliest days on a kind of openness and a kind of sharing, and it shows up up and down the stack. So, at the very highest level, you’ve got Berners-Lee’s original vision of all the world’s knowledge linked. And then you have these tiny, seemingly minor details that ramify in enormous ways.

So, Marc Andreessen, who writes the first graphic user interface browser called Mosaic back in the early 90s, there’s a button in Mosaic that says view source, right, show me how the HTML code makes this page look the way it does. And it’s just a debugging function, right? It’s for techies, nobody wants to look at that junk. But, Andreessen didn’t turn it off.

It would have been the easiest thing in the world to say, OK, now we’re shipping this to Ma and Pa Kettle. They’re not going to view source so we’re just going to turn view source off in everybody’s browser and then if you want the specific techie version, you can request it, or something like that. They didn’t do that. So, everyone’s browser to this day has view source enabled. Almost no one uses is, but if a kid wants to see how a web page is made, they can view source and get some sense of the structure.

And that reflexive notion of openness isn’t just about the technology, it’s also about the culture. Brian Behlendorf for the Apache Web Server, exactly the same thing. And so you have these people who are working on tools but also embedding cultural beliefs about sharing into the design of those tools. And you can’t really, I think, tease apart those two things. The Web started with this feedback loop of technological and cultural openness and that continues to this day.

Nora: I mean I was thinking about that, you know the phrase of Marshall McLuhan, “We shape our tools and afterward our tools shape us,” which does make you wonder how once we’re swimming in this sea which is…

Clay: He missed step three, though.

Nora: What’s step three?

Clay: After that we shape our tools the way they’ve shaped us, which is to say, at least the simple reading of that phrase is there’s a kind of deterministic, the medium is the message, we occupy the tools we create. But, in fact, we’ve seen over and over again that user innovation happens within a given set of tools, and that the designers of those tools will watch what the user is doing and alter the tools accordingly. And that becomes yet a new environment for the users to play around in. And it’s that feedback loop, I mean McLuhan’s phrase encompasses both the sense of the feedback loop, but it conveys, I think, a more static sensibility about the relationship between user and media than the one we’re witnessing today.

Nora: I mean, I am really interested in this idea of how the users shape the tools because it seems self?evident that that’s happening, and yet we have this whole ideology about the inventor in the lab by themselves who hands them down like God giving Moses the tablets, or whatever. Do you think there’s something about the environment that we’re living in now though that makes it easier or, yeah, easier for people to shape the tools that they use?

Clay: Yes, absolutely. There are several things all going on at the same time. The first is the one we talked about, which is just it’s much cheaper to do any kind of playing around. The second is that we’re living through, in the media environment, an enormous shock of inclusion where the public, previously locked out of participating in the media except as consumers, now anybody can say anything, anytime they want, and frequently does as you can tell if you wander around LiveJournal on any given afternoon. And that shock of inclusion isn’t just about individual freedom, it’s also that social visibility. It’s very easy to see what people are doing and to start riffing on it.

So, the bestselling T-shirt on Amazon, the Three WolfMoon T-shirt, got to be the bestselling T-shirt on the Amazon, not because it’s such a great shirt – it’s the kind of shirt of a 15-year-old boy would wear if he wanted to look tough – but because there’s a hilarious user review that then kicked off this whole drama of fake user reviews; 200 people pile in.

So, the fact that people can see other playing that kind of game over there, and I’m going to join it, and that whole group can create something that’s even more visible, that means that… When you get things like Friendster, the original large scale social network service, people just started playing around, like they started setting up Friendster profiles for evil himself or the city of San Francisco, or whatever; so you and I would be connected through our mutual friend evil himself.

And Friendster saw this kind of user-generated playfulness, and they said, ‘This is wrong; these people must be stopped,’ and they, by preventing the users from engaging in the kind of playfulness that helps tools adapt, they drew Friendster, first into a kind of dead end and finally into the brickwall at the end of that dead end.

Facebook, I think, and MySpace learned that lesson by having a much higher degree of user freedom to kind of configure and play in that environment. That’s probably the biggest scale example we have right now ? the difference between discouraging and encouraging the users from experimenting with the tools in a way that changed the tools in a way that changed the users.

Nora: I mean, this is a huge cultural change for anybody who runs the business to get their heads around having that attitude towards their users. So, what would you say to encourage somebody to say, OK, yeah, that’s actually a good thing if your users are mucking around the stuff and doing things that you didn’t imagine them to be doing?

Clay: Yeah. Well, so it’s easy to make that case to owners of platforms. For instance Flickr, founded in Vancouver – which I was on the advisory board of back in the day – set themselves up to share photos, but then quickly you started getting groups of people coming around. And so there was one early group, square in circle, rather circle in square. They’d try and take a square photo, camera phone style, with a perfect circle right in the middle it, touching all four sides. This was the game, and you get thousands of these photos. And Stewart and Caterina, the founders, were smart enough to realize “This is useful,” because it drives adoption, it drives use, it make it more interesting.

So, if you’re running a platform, if you are running Flickr, or Delicious, or YouTube, or whatever, user playfulness is great. The harder cases for businesses that are just participating in the media environment, and there it really is more of a mixed bag. Because a lot of what the users are going to say about a business is to complain about the service, or to wish they had something different, or better, or whatever.

And the trick, I think, is to view what the users are saying about the service, to take that away from the marketing and PR departments, and send it instead to the product department. Which is to say, you start viewing user feedback not as some negative in a media landscape you used to control, but rather as information about how to do it better next time. Dell famously has opened itself up to input by users.

I recently bought a netbook, and I bought one from Hewlett Packard because on their site, they have user reviews. And the users are saying “This one is good, and that one is not good.” And that degree of transparency made me think “OK, I’m going to buy the one that these users are recommending that I buy.” Because if HP is willing to expose itself in my sight to this kind of feedback, that suggests to me a long term commitment to quality.

For businesses that thrived on being able to control the media landscape down to the last sort of jot and tittle, principally service agencies, advertising agencies, PR agencies, a lot of this stuff is nothing but bad news, because it’s uncontrollable, it’s unfilterable, it’s unpredictable, and so you move from a predictive and prescriptive model to a reactive one. And I think every communications media marketing agency in the world is dealing with that, in one way or another. And the good ones are figuring out interesting new things to do. But there, the more a business relied on the consumer not being able to talk back, the worse news the Internet is.

Nora: I’d like to pull in some of the ideas form your book “Here Comes Everybody.” It seems like when we are talking about the change in how innovation is happening, it’s this example of a larger shift from the old idea of a company hiring an R&D department, and sort of assuming the cost of innovation and so on, and putting it out in the world. To something that’s a lot more loose and collaborative. Can you sketch out that shift a little bit?

Clay: Sure. Well, actually the interesting thing about the collaborative piece is that the business model, the sort of classic enterprise model of collaboration, I put together a design or an engineer and a project manager, and so forth, and they go off in a room and work together. The surprising thing about web scale collaboration, about the design or creation of large artifacts, e.g. Wikipedia, or the Linux operating system, the Apache Web Sever, these kinds of things, is how successful they are at reducing the amount of collaboration. One of the really big changes from enterprise models, to very large models of work is that you can take in proposed changes or improvements to a product, like Wikipedia, in tiny, tiny, increments. I can go onto Wikipedia, and I can fix a typo. A single typo on a single page, and never touch the thing again. And nobody has to cut me a check, and nobody has to check my credentials, or any of the rest of it. And that small granule of work at very, very low cost of integration is absolutely essential. Because if I actually had to collaborate in any serious way, people know who I am, we get to know each other, I would never go around fixing typos.

And so for very large scale creation, what you get is this funny model of co-creation without collaboration. There are tight collaborate groups, highly, highly active, sort of at the center of things, managing the process, right? The people who are managing what finally goes into the Linux kernel are very, very careful about what they do and don’t accept.

But, the people at the edges just messing around with the code to see they’re doing, no one is keeping track of those people. They don’t have to. And interestingly, the design of what’s called the Source Code Control System, the way the Linux project keeps track of all of the proposed changes, is specifically designed to maximize the freedom and minimize the oversight of the lowest level participant – which is the inverse of the business model. When you come in and you just got there, they minimize your freedom. All you can do is fill out this one form this way. And then after six mouths we will see how you are, and we’ll give you a little more freedom. Here, it’s like “Look, check out the whole kernel, do whatever you like. We don’t care.”

The chances that you’re going to do anything important or interesting with it are effectively nil. But, they’re not absolutely nil. And so we’ll hold open the possibility that somebody we never heard of puts in a security patch that actually turns out to be really important. And if we never hear from that person again, that’s great. And that’s a model that’s so different from the enterprise, that there are still people writing today as if Linux or Wikipedia should either not exist, or will go away tomorrow because they can’t understand how it’s happening, even though these things have happened for a decade, two decades now.

Nora: This may answer this question I’m going to ask, but I twittered earlier today that I was going to be interviewing you. And Elizabeth Wellburn talked about research that pointed out problems with group decision making – the thought that people tend to engage in groupthink, and so on. And she wanted to know, not because she’s critical of Web 2.0 or anything like that…

Clay: She should be.

Nora: … but she wonders what it is about the character of the web now, that makes things different?

Clay: Well, I think it’s the thing we were just talking about, which is that very often, the groups that we’d be afraid of groupthink in, are actually not collaborating. Groupthink comes about when people commit themselves to the identity of the group as a whole, and become participating members. Now, culture is a really, really important human characteristic. Given how flexible we are as creatures, we need some way of reducing all the possible ways we can relate to one another, in order that we can have any kind of sensible interactions. And so different cultures arise, all the way from macro scale cultures, you know “The West,” all the way down to “Right here’s the culture of these kids at this high school, and they have their in jokes, and their selected gathering spots, and so forth.”

For all the good that culture does in terms of synchronizing people however, it is also the producer of this kind of groupthink. And Cass Sunstein – who wrote a book called “Infotopia,” which covers a lot, does an amazing literature view on the groupthink stuff – says that even worse than groupthink, is that when groups get it wrong, they are likelier to persist in wrong beliefs and to believe them more strongly at any given moment, than individuals with the same wrong beliefs.

So, what’s different is that you’ve actually got systems where if I disagree with the entirety of the Linux developing team, or the Apache developing on say one of these open source projects, I can take the entire code base and go off and do my own thing. No one can prevent me from doing what’s called forking, which is to say to create a completely alternate project. And as a result of forking, it’s as if I didn’t agree with the polices of my country, I could form my own country of one. And that inability to force individuals, in at least some of these situations, to force individuals to go along with the will of the majority makes the arguments much more contentious, but it also preserves them as technical meritocracies.

And that’s a big change. It used to be that groups had to coordinate themselves and assume a group identity and culture before they could get any work done. But now, if you can get work done with thousands or millions of people who you don’t even know, the risks of groupthink are smaller than they were.

Nora: So, how big a change do you think this is culturally?

Clay: Wow, that is a huge question. I used to be for, you know, space of 18 months, two years, I used to be utopian, which is, you know, I got hold of the net in the early 90s, and I thought oh, this is unbelievable, you know, this is the dawn of a kind of digital Age of Aquarius. Then I turned into a mere optimist, which is to say I started to see that there were downsides, and I thought that on balance, the upsides would outweigh the downsides.

But recently, I’ve really had to sit down and say, I am a revolutionary, which is to say, I think that this change is so large that its ultimate effects cannot be predicted from the middle of the maelstrom. So, I can sketch out a scenario in which every hierarchically managed organization, whether for?profit, not?for?profit, government, private sector, whatever, is transformed because the thing that holds large organizations together is communications, and when there’s enough of a shock of inclusion in the communications environment, everything changes.

I can also sketch out a scenario in which there are kind of two worlds that exist in parallel. One is the world of assumed openness, and that world grows and is creative in all the ways we’ve seen, but that it ultimately reaches some accommodation with the hierarchically managed world and the for?profit world and so forth. And we then know the outlines of those two countries.

And I don’t feel confident at this juncture, or rather, when I catch myself extrapolating after that first world, everything changes, I realize that what I’m doing is extrapolating present trends infinitely into the future, which is a classic mistake of prediction. So, I will say this: In the places where this is a change, this is a huge change. The media environment plainly is getting it first, as it would, being based most directly on the communications capabilities of a society, but I see the changes showing up lots of other places.

On the other hand, we have lots of examples of institutions that continue to exist and find new roles for themselves even in the face of enormous technological change. So, it’s going to be a very big deal in at least some places. It will change things in a lot of places. And I can’t, I just, I can’t extrapolate either of those thoughts, both of which I believe completely, to a stable end state that I feel like we can predict from 2009.
If you had gone to Germany in the mid-1470s and said, let’s see what this printing press is doing, right, you would miss novels, you would miss newspapers, you would miss the rise of scientific publication, you would miss Martin Luther’s “95 Theses,” you would miss the Venetian publishing industry. So many of the changes brought by the kind of abundance created by the printing press were in the second 50 years of its existence, if not the second century of its existence, that I think that over-extrapolating from current trends would leave us in the same position as if we tried to do the printing press in 1473.

Nora: Are there things about this sort of loose way of innovating and developing that concern you, that you think we should be cautious about or that we need to be aware of addressing?

Clay: You know, in a way, those break down to two different questions. I don’t think, I mean, the people who are concerned about the Internet upending culture as we’ve known it are in general, as people who make those kinds of predictions always are, they are in general basically correct, right? When the people said, oh my goodness, you’re printing Bibles in French and Italian? This is going to be terrible! This is going to wreck Western civilization as we know it. And that’s, indeed, exactly what happened. It wrecked Western civilization as we know it. There’s a permanent tension between freedom and quality in the media landscape, and that’s just the great, you know, the two poles of the media debate. And any time, any time there’s a shock of inclusion, the dial tips towards freedom, average quality falls, and then you get all this experimentation.

So, I don’t think that there’s anything, I mean, you know, I grew up in the 70s, right? I remember what culture was like before there was any digital expression in the hands of the public, and it’s not a culture that I would particularly say, that was the apotheosis of the human race, we definitely ought to be preserving that.

Nora: There was “Fantasy Island.” Come on, you know?

Clay: There was “Fantasy Island,” exactly, but this is the thing. In a way, mass media achieved its high water mark in the period just before the Internet, in part because the audience was at the scale it was at, but there was no alternative, right? There was nothing we could do except, you know, if we wanted media, we had to consume it as it was being produced, and that’s broken. So, there’s nothing I’d say, oh, we must form a cordon around, as some people are saying, oh, we must form a cordon around newspapers, or you have to say, newspapers in their traditional form, or we have to preserve the university in its traditional form, or whatever. I don’t. I’m much more worried about the problems that are native to group collaboration.

Nora: Right.

Clay: Not the part of the old society that breaks, which, you know, that happens. Here’s the thing that worries me. Openness creates experimentation, again, with cheap enough technology, openness creates experimentation. Experiments create value. Value creates an incentive. But, the incentive, if all the incentive tied back to was more openness, right? We have this lovely, virtuous feedback loop. And in the early days of any medium, it looks like that’s what we get.

But, incentive is actually completely value-neutral. The spammers are responding to the incentives of email just as surely as you and I are. People trying to spam Twitter, people trying to game voting systems online, certainly the people, you know, identity fraud, credit card theft, all of these things are responses to openness just as surely as people blogging about their life’s experiences or Twittering to their friends or updating their Facebook status is.

And one of the things we’ve seen, particularly with Wikipedia, is that past a certain point, openness becomes indefensible. The only response is to partially close down. And so, the history of the last three or four years of Wikipedia’s governance structure has been a very targeted attempt to find those places where the incentives are the most deleterious to the enterprise as a whole, they tend to be around biographies of living persons, and to do things like having semi-protected status, locking articles, flagging revisions, a whole variety of techniques for saying, it’s the encyclopedia anybody can edit but either, and not you, or, and not now.

What we don’t know is, is this an arms race where it’s just fought to a kind of stalemate that, you know, each side kind of ups the ante but basically some new steady state is reached; or are we in the middle of a process of progressive decay, where at a certain point, the number of people who want something like Wikipedia to fail essentially lock out its initial promise from most of its current participants – I hope, because it’s something I want to believe is true about the universe, frankly, this is now just deep desire, it has nothing to do with analytics – I hope that the kind of bargains we’ve seen in the design of constitutional democracies can start to enter this sphere, which is to say, look, it’s a group of people who’ve got conflicting needs, and here’s how we’re going to balance it out.

But, I also recognize that there are systems, and we have examples in the past, like CommuniTree, this bulletin board that was overrun by high school kids and eventually just pulled the plug, they just shut themselves down because they couldn’t withstand the onslaught, that there are models where it’s not like the design of the Constitution, it’s like the sack of Rome…
[laughter]

Nora: Right.

Clay: … Where the people coming in intent on destroying can simply have the upper hand, have or gain the upper hand. And we’ve seen more models that are like constitutional design than like the sack of Rome, but you can’t write off the desire of people to destroy things they don’t like or don’t trust, and that’s still an open possibility.

Nora: Clay Shirky, thanks so much.

Clay: Thank you. Thanks for having me, Nora.

Nora: That was my full, unedited interview with author Clay Shirky. An edited version of this interview will air on an upcoming episode of Spark. If you like hearing these kinds of full, uncut interviews, you can subscribe to the Spark Plus podcast. You’ll get all the regular weekly episodes of Spark plus bonus blog?only audio. It’s totally free, really easy, so check it out at our website, cbc.ca/spark. And thanks for listening.