2014 Hungarian Internet tax protests

Transcription

JONATHAN: Good morning, and welcome to Talks at Google in Cambridge, Massachusetts. Today it's my great pleasure to introduce Ethan Zuckerman. Ethan directs the Center for Civic Media at MIT, and he teaches at MIT's Media Lab. He's the co-founder of the global blogging community Global Voices, and he works with social change nonprofit organizations around the world. Much of his research centers on quantitative analysis of media, computational techniques to determine who, what, and where feature in traditional and participatory media. He's here today to discuss his book, "Rewire, Digital Cosmopolitans in the Age of Connection," now in paperback, which you've never seen before today, so small secondary thing, the book which "Publisher's Weekly" called "a fascinating and powerful reflection on what it means to be a citizen of the world in the internet age." Please join me in welcoming Ethan Zuckerman. [APPLAUSE] ETHAN ZUCKERMAN: Thanks so much, Jonathan. And thanks, everyone, for coming out. I really appreciate it. Obviously it's an arduous trek from across the street to be with you. So as you guys know, I'm sure, writing is a bit of an isolated process. And so you go into this cave for 12 to 18 months, and you shut everything out of your head when you try to get the book out. And then you put this book out, and you hope that the book is still relevant. You hope that the ideas that you put forward still help explain the world. So I was really gratified shortly after the book came out to get a story that sort of summed up for me the core point that I was trying to make. And as it happens, it's a Google story. So I didn't make this up especially for you guys. This is the story that I've been trying to tell people to get a sense for the phenomenon that I'm talking about. But it centers on this guy Hugo Barra, who's quite senior at Google, working on the Nexus product, very much the person who would stand up and show everyone where this remarkable phone was going. And you guys may remember about 18 months ago he announced that he was going to leave Google. And specifically he was leaving Google for Xiaomi. And Xiaomi was a company that at that point most Americans, including a lot of Americans in the tech press, knew very, very little about. And so there was something of a buzz in the tech press around the idea that someone would leave this incredibly prestigious job at Google and go work for a Chinese mobile phone manufacturer. And so the tech press did what they're so good at doing. They actually dug deep down. They found the story behind the story. And they started publishing stories explaining that what had actually happened was that Hugo Barra had found himself involved in a complicated love triangle with Sergey Brin. And of course, if you had been dating a Googler, and that Googler ended up dating Sergey Brin, there was only one logical thing you could do. You had to leave the United States and go to a place where Google's drones couldn't reach. And you were suddenly outside of the orbit of where Google had any influence. And this was obviously the only thing that could explain Barra would give up a dream job and find himself going to work for a Chinese mobile phone manufacturer. And this got reported with a fair amount of seriousness. You can look this up and you can find the stories of this triad. And so Barra took to the stage at LeWeb last winter, and he said, guys, you really just don't get this. If you want to be working on tech right now, particularly if you want to be working on mobile tech, it's an amazing time to be in China. And it's not just an amazing time to be in China because you're looking at 500 million people on the internet, you're looking at this massive emergent middle class. It's not just a great time to be there because Xiaomi is actually a pretty interesting company and doing innovative things on top of the Android operating system and building really beautiful phones and running them in small runs on a unique business model. It's an interesting time to be in China because the Chinese internet does not look a whole lot like the US internet. And in fact, a lot of the tools that have become dominant in the Chinese internet actually don't have very neat US parallels. We talk often about Weibo as being sort of Twitter on steroids. It's actually significantly more interesting in political terms. It's really become sort of the space for public discourse. WeChat with the circles model has done something that even Google has actually had trouble getting across with Plus, but a really different model for communication within small groups that's worked very well. Alipay has brought digital cash into ordinary, everyday use in a way that, again, isn't all that common at this point in the US or in Europe. As much as we may be impressed by what our friends over Amazon are doing as far as delivery, JD.com putting things on people's desks within four hours of ordering is really an order of magnitude beyond this. And so I don't expect these stories to be particularly surprising in a room like this. But generally what I'm going out and talking about what's going on on the Chinese internet, I get a whole lot of bafflement, because there's this sense in which we expect the Chinese internet to be boring and stultified and censored and non-generative. And in fact, in many ways it's one of the most interesting spaces that's going on right now. And so I'm not actually here to talk about the Chinese internet and why we should take the Chinese internet seriously, although it's a fun topic and an interesting one. What I really want to talk about is this broader notion that I've been wrestling with in this book, which is this possibility that atoms may, under circumstances, be more mobile than bits. We're actually remarkably connected to China in terms of getting stuff in and out of China. We've all gotten our heads used to the idea that we are engaged in supply chains that are deeply global in form. And you start thinking about what's involved with building the laptop that I'm presenting from, and you're involving 50 or 60 countries. You've got container ships moving between all of them. And this sort of miracle of a designed-in-California, made-in-China, delivered-directly-to-me product gives you a sense of just how connected these sort of networks have been. What's interesting is that I'm making the argument in this book that the digital networks that tie us together don't actually work as effectively. And the way to get this is you can't think about these networks in terms of potential. You have to think about them in terms of flow. You can't just think about the idea that in theory we're capable of getting all sorts of information out of China or all sorts of information out of Sub-Saharan Africa. What I'm really asking is, where does that information flow? It's really hard to map flow. We get it a little bit when we look at a map like a visualization of traffic. And we look at the city of Cambridge, but we also look at where you want to drive and where you don't want to drive at that point. This is a map of air routes that actually does a nice job of flow. You get out of it is a sense that you can probably get to any point in the world. But if you start looking for where it's particularly thick, where it's particularly connected, you start getting very good indicators of where people actually go in practice. But even a map like this tends to disguise a little bit what's going on. What actually happens when you look at air traffic is that about 90% of it is domestic. It's actually almost a rounding error, the international traffic. We think about it when we start thinking about how we interact with people from other countries. We think about it when we think about how disease spreads. We think about it when we think about just-in-time delivery. But in fact, the vast majority of this ends up being domestic. And what I'm finding is that the same thing ends up happening when we start dealing with flows of information. By the way, I have to think that this is one of the great intern projects of all time, right? This guy Paul Butler goes over to Facebook, says I'd like to map Facebook, gets enough data that he's able to do this visualization. You can find Zuckerberg pointing to this map every single time he talks about Facebook's global ambitions. So just in terms of like constructing a map as rhetorical document, I have to think that this is sort of a fun thing. But what this map very clearly wants to tell you is that Facebook is this incredibly global network. What it actually doesn't do a very good job of telling you is how much of that traffic on that network is people who are literally in the same town, in the same city, in the same country. We've got this illusion, this vision of the cosmopolitan by looking at a network that theoretically connects us all. But when we start asking this question-- where are we actually getting our information?-- what are we actually paying attention to?-- are we hearing the story about Xiaomi?-- are we hearing about Ebola before it becomes a crisis?-- in many cases, we're actually dealing with a remarkably disconnected world. And for me, what I study much of the time is what we pay attention to in news media. And so this is a tool. This is actually publicly available if you want to play around with it. We are scraping 50,000 media sources in the US and then doing visualizations day to day based on our own geocoding coming out of it just to get a sense of what we're paying attention to and the incredible imbalance between domestic news and international news. In some senses it makes sense. Everybody pays more attention to what's going on domestically than they do internationally. But the extent to which, at a moment where it's almost possible to get information from anywhere, we have a tendency to be overwhelmed by domestic information, is to me sort of a surprising one. But I want to talk about why I'm interested in this. And for me, it has a lot to do with my own history. I got involved with building internet-based businesses in 1994. I was the founding CTO for a company called Tripod.com, which did homepage hosting. It get swallowed by Lycos. like: Lycos got sold off into a million different parts. But a lot of the things that people working on user-generated content businesses are dealing with now we were dealing with in the late 1990s. I came out of that and said, well, this is really interesting. We built a really big company that was the eighth largest site on the web for a little while. We built it in a town of 8,000 people in northwestern Massachusetts. I wonder if we could do it in West Africa. Because of course that's a logical next step. It's western Mass, West Africa. So I started a group that invited technology volunteers to come over for three to six months at a time and work in start-up businesses in West Africa and try to do tech transfer and help businesses get off the ground. And believe it or not, this is in 1999 one of the hottest businesses on the continent. This is Joy Online, which has turned into Ghana's main international portal. If you're a Ghanaian living anywhere else in the world, this is your source of news and commerce and revenue. This is your connection to the rest of the world. And so I was hanging out in Accra, Ghana with these businesses and seeing amazing things, people building really fascinating businesses, high levels of technical sophistication, wondering why the world hadn't figured out that Ghana was going to be a great destination for outsourcing. And then something really interesting happened. Ghana had an election, a really good election in 2000, elected this guy John Kufuor. And Kufuor was an interesting guy. He came from the opposition party, the party that had been in power for more than 20 years, gave up power voluntarily. This guy got elected in an election that was universally seen as free and fair. And as you can tell, it dominated all the news in the newspapers. So we, just to remind you in 2000, were having a bit of an iffy election here, Bush v. Gore, didn't go actually all that smoothly. It was actually going so badly here that Ghana offered election observers to come down to Florida to try to help them run this election. And unfortunately, this incredibly good news story out of Africa ended up getting something like 200 words in the "New York Times." And I was looking at this and essentially saying, there is no way that I can do what I'm doing. There's no way that I can figure out how to get American companies to take Sub-Saharan Africa seriously unless we start looking at this question of what we pay attention to. And so I started a scholarly career looking at this question of, how do we pay attention to the rest of the world? And what's been really weird about this is there's a huge shrinkage in international attention. If you looked at international newscasting in the 1970s, at least one out of three stories was an international story. Now, there's some reasons for this, right? This is the height of the Vietnam War, a lot of this is America's involvement in Vietnam. But there was a lot more international attention at that particular moment in time. And this erosion in broadcast coverage is due to a lot of different things. It's due to cost. It's a lot cheaper to cover domestic events. To a funny extent, politicians sort of cover themselves, as do celebrities, whereas wars actually require you get out on the ground. This has happened not just in broadcast. We've replicated this finding in print in the UK and in the US. We've seen a 45% decrease in coverage of international news stories between the 1970s to the last decade. And when you think about it, this goes completely counter to what we would expect from the technology. We would expect that in the 1960s and 1970s when it was incredibly challenging to go and get these stories-- you had to go to Vietnam, shoot on film, put the film in a canister, ship it across an ocean, develop it, and then put it up for broadcast-- that we would have a lot more information coming out at a moment where you can put a crew on the ground with a satellite phone, with a camera, get footage from Haiti within 12 hours of disaster happening. But most the time, that isn't actually what happens. And one of the big reasons why it doesn't happen is that we have a surprisingly light appetite for news. This is work that I did just scraping off of Ad Planner, back when you guys left it open and I could scrape it, just trying to get a sense for what was the appetite for domestic versus international news in the largest internet markets, so just looking at how much attention was paid to various different news sites. Turns out that the US is actually quite cosmopolitan in our news attitudes. We pay a lot of attention to the BBC. We have a very large Indian expatriate population that religiously reads "Times of India." But even with that much internationalism, we're at roughly 94% domestic sources. So consider "The Guardian," consider the BBC, consider all of that. We have a strong preference for the domestic. It's much, much stronger in most other nations, this sort of promise that we could look at the internet, and I could read the "Times of India," I could read various different Australian or Chinese papers depending on my language skills. Most of the time this is not where we end up spending our attention. So we end up actually very dependent on the coverage that comes out of mainstream media. And that coverage has had biases that have stuck with us for a very, very long time. I've been mapping the "New York Times" for about 10 years now. I actually have to mark my maps very carefully, because I can never tell by looking at them, because the pattern doesn't really change. We're almost always under-covering Sub-Saharan Africa. We're very rarely covering Central Asia. We tend to have a very, very strong interest in North America, Western Europe, and increasingly in India and China. But of course, now we're in a digital age, right? So we've got infinite shelf space. We're not constrained by what comes out on the paper "New York Times." Now you start seeing something like the Huffington Post trying very, very hard to internationalize. And what you'll see looking between the two is you actually end up with a much paler map. You end up with significantly less international exposure as you start moving from the Times into HuffPo. You actually have a much stronger concentration of stories coming out of the United States, stories coming out of Europe. And so this is why, in many cases, people sort of look and say, well, wait a second. OK, so we still have strong biases within existing mainstream media. As we're moving into digital media, we're ending up with possibly an even more Ameri-centric media. What about this graph? Aren't we going to get interesting international news through our friends? As social kicks in, aren't we going to find out about really important stories because our friends in other countries know about them? So I had the great good fortune to get to interrogate this data a little bit with a guy named Johann Ugander, who wrote this very, very helpful paper on the anatomy of the Facebook social graph. And what this paper suggests is actually a pretty surprising to folks like me who study international contact. It's suggested that 15% of the relationships on Facebook cross international boundaries. And I found myself looking at this and going, that's amazing. That's so much higher than we would get in any other sort of statistic. And so we hammered at it a little bit. And as we hammered at it, we discovered, first of all, that a lot of this was coding error. It's actually closer to 10% that was actually international. And that 10% international is actually highly clustered. To the extent that Scandinavians hang out with other people, they hang out with other Scandinavians. To the extent that people in South America hang out with other people online, it's other people in South America. And that's more or less what this graph shows us. But what we found that was most remarkable in this is that the average person has more like 2% to 3% international ties rather than 10% to 12% international ties. And the reason for this is if you are a Norwegian living and working in the United States, you show up as having 80% or 90% international ties within this data set, because you're using something like Facebook to stay in touch with everyone at home. So you may be what we refer to as a bridge figure. You may be someone who is able to bring interesting stories from Norway and bring it into a US audience if that US audience is capable of paying attention to it. But you are a very, very small subset of the population. And the rest of that population is dealing with a social network delivering them stories and delivering them information that's really governed by the phenomenon of homophily. So homophily was a word that I didn't know before I sat down to start writing this book. When I found myself dealing with it, it unlocked a whole trove of sociological literature for me. Turns out that homophily, which is basically the tendency of birds of a feather to flock together, is probably the best documented finding in contemporary sociology. People are incredibly good at self-selecting into comfortable tribes of one fashion or another. We see evidence that in neighborhoods, people will move in next to people of the same race, next to people of similar socioeconomic status. You see the same thing when you walk into a cafeteria and people don't know each other very well. They tend to cluster by visible characteristics like race. But if you actually interview, you'll find out that their unconsciously very good at clustering themselves by religion, by socioeconomic status. One of my favorite papers in this space looks at students walking into a computer lab and finds that students end up sitting next to people who have similar facial features. If you wear glasses, you're likely to sit down next to someone who has glasses. If you have long hair, you sit down next to someone who has long hair. And when you interview people after the fact to try to explain this, they say, they looked friendly. We're really good at trying to find our tribe. And what's interesting about this is that this tendency to find our tribe, I think, is playing out online as well. When we're dealing with curated media, like the "New York Times," we pay more attention to our tribe than to the other tribes. We pay more attention to North America and to Europe than we do to anywhere else. And in some ways, the further we get in terms of cultural distance, when we're talking about something like Sub-Saharan Africa, you're going to get less and less information. As we move into an era of choice, where I, as the user, say, this is what I want to know about, when I search for something, when I customize my news, anything within that paradigm basically now depends on my ability to be conscious of my own choices. And unconsciously, I'm likely to pay more attention to people who are like me. And so when I think about, give me what's important, give me what I need to know about in the world, I am likely to have a somewhat biased approach to it based on my own natural tendencies to pay attention to my tribe. And now as we start moving into an era of social media, it's interesting that this is going to play itself out even further. Because most of us have pretty homophilous social networks. We have a strong bias towards people who have the same ideology, who have the same religion, who have the same basic background. And if you think about the fact that a network like Facebook starts by literally trying to recreate your high school class and put you back into your town of origin and connect you with all those people, you have a very, very strong homophilous bias to what sort of information comes in through there. Now the argument that I'm making, by the way, is actually pretty different from Eli Parser's argument in "The Filter Bubble." he's saying it's all about the algorithms. They're lying to you. They're just trying to give you what you want. I'm actually saying the problem is much worse than that. The simple problem is, because we have this very strong natural tendency to pay attention to those who have a lot in common with us, even a completely unbiased, a completely un-logarithmically controlled network, would still fall into something of this homophily trap that I'm talking about. So why should we care about this? I think the big reason for me to care about this is that the really big, really interesting problems probably don't get solved locally. It's very hard to find someone who believes that addressing climate change in a serious way is going to be something that the US does by itself. There may be technological innovation. There may be major breakthroughs. But at the end of the day, this challenge of trying to figure out how China, how India, how Myanmar, how Nigeria industrialize without having the same carbon footprint that we have is an enormous international problem that's actually going to require us to have conversations with the emerging middle class in those countries and try to figure out how to both meet their aspirations and also try to minimize the impact. But I think in some ways, we actually have gotten a much better example just in trying to figure out what's happened with Ebola over the last couple of months. If you spend as much time reading about West Africa as I do-- and I do it, again, because of my own homophilous tendencies, right? I hang out with a lot West Africans. I read West African news. We knew that this outbreak was something particularly scary and particularly serious in mid-March. But what's interesting about it is that there was almost no attention to it. This is part of the same suite of tools that I was showing before. This is the media cloud set of tools. And what I'm looking at here is the 25 largest-audience US News sources. And what I'm looking at here is a search for Ebola and Guinea, which is that orange line there. And you'll see that back in March, back in April, we were talking a little bit about Ebola in Guinea. We have almost nothing until the crisis really starts getting out of control in Guinea. People are really, really worried that large chunks of the country are going to get knocked out. And we're talking about this in July. Now, the blue line, by contrast, shows you how we normally pay attention to this. That's the search string "Ebola and Texas." So as soon as this hits national borders, as soon as this hits people like us, people that we pay attention to, people that we care about, we have this massive surge in attention. The thing is, it's too late. This is now actually a very serious epidemic. It's going to affect tens of thousands of people. It's possible that West Africa is going to end up being a reservoir for this disease for some significant time to come. We could have had a great deal of impact on this had we jumped into it very early on. But it's incredibly hard for epidemics to get onto the news agenda. They're not like earthquakes, where happens one day, it's an act of God, and everyone sort of rallies together, and let's go save Haiti. It's the kind of thing that you actually have to pay attention to day after day and get a sense of how serious it is. And there's a great story on "Planet Money" about why it has proved almost impossible to raise money to fight the Ebola epidemic. And instead, what we have are people and in essentially saying, let's just shut our doors. I think Ebola is actually a lovely example of why this just doesn't make any sense. We know that people are going to keep going back and forth. We know that people who are caring for their families in West Africa are going to find ways to cross borders. They're going to end up doing it perhaps by lying. And if they lie about where they've been, we're not going to be able to monitor. We're not going to be able to take people's temperatures twice a day for 21 days. We're also going to lose the ability to send the medical personnel who want to go over and actually help with this and be able to go back and forth. This is one of these situations where this tenancy to say, we should just withdraw from the world, when you start playing it through, really doesn't make a whole lot of sense. This is a situation that in a connected world we all end up having some responsibility for. But I think the reason to care about this stuff isn't purely negative. And where I end up in the book is talking about affirmative reasons why you might want to be paying more attention to the rest of the world. And a lot of this leans on this guy Ron Burt at University of Chicago. And Ron left U Chicago to go work for Raytheon and to pretty much conduct just one of the greatest social science experiments of all time. He got himself hired as a senior vice president at Raytheon just as Raytheon was trying to integrate five new companies that they'd found the money to pay for after inventing the Patriot missile. And so you've got a giant defense contractor trying to figure out its own internal structure, how does it work and how do ideas spread within the company. So he takes one department. He takes Internal Purchasing. And he interviews everybody in the entire company, more than 1,000 people, and says, write down your best idea. How would you make Raytheon better? People write them up. He anonymizes them. He puts them in front of senior management. Senior management ranks all of these ideas, identifies the people within the organization who have the best ideas. And he tries to figure out what this might correlate to. You would hope that this would correlate to your highest ranking people, your most experienced people. The answer is no, actually-- fairly loose correlation between whether you've been with Raytheon for a while, whether you're in a senior management position, whether you have a good idea or not. It correlates incredibly closely to the sociogram of the organization. If you are a bridge between structural holes, if you are someone who travels between different offices, if you are the person in the Hong Kong office who is in touch every day with the Honolulu office, if you are someone who has moved throughout the company and so you have ties in a lot of different offices, you are at, as Bert puts it, an alarmingly high risk of having good ideas. He's able to look at this whole organization and essentially say, we think the best ideas, as judged by management, are coming from the people who have the access to the most diverse ways of looking at a problem. And he ends up coming out of this with a really interesting theory. He says, creativity is an import-export business, that we think about genius in terms of being Picasso and being able to see the world in an entirely different way. But what genius is most of the time is taking something that's incredibly ordinary in another setting and putting it into a totally different setting. And as it turns out, I'm not just randomly juxtaposing Picasso and a mask from Benin. Picasso started collecting West African artwork shortly before he invented cubism. And if you think about Cubism, what it largely is, is about taking these sort of geometric shapes that come out of African mask-making and bring it into high art painting. Burt suggests that one of our ways to try to be Picasso is to try to figure out how to diversify those information networks. How do we amplify our cognitive diversity? How do we make sure that we are interacting with enough different people who see a problem in enough different ways that we're able to actually approach it from multiple different approaches? So I've been working on this question, this question of, how would we build an internet that's good for looking at the entire world, for the last 10 years or so. And I'm really bad at it. And one of the reasons why it's fun to come to a place like this is because the things that I've done so far have been so remarkably, dramatically unsuccessful. This is the dramatically unsuccessful project I've been working on for 10 years. This is a community called Global Voices. It's about 1,200 volunteers in about 130 countries. They put out a daily newsletter of what's going on in social media all over the world. And so if people right now in Djibouti are talking about whether the overthrow of the government in Burkina Faso provides a model for Djiboutians to rise up against their government, we've got it. And that happens to be the lead story this morning. If people are talking about pop music in Ukraine, and that was a lead story the other day, that's what we've got out there. And our whole idea behind this was we'd go after the supply. We would basically say, we know that we've shrunk how much time we're spending on television or in the newspaper international news. Let's go after the digital. We will look for people who are great bridge figure. These are people who are within these countries but also speak English. They can help you figure out what's actually going on, what might you want to pay attention to. And while I think the model is really good, what we never thought about was demand. And we've never managed to get this product more than about a million people a month. There's about a million people who are actually quite passionate about it. They spend a lot of time on it. They're really excited about it. But most people either don't agree or haven't gotten the message that you would want to think about how you get more diversity of what you're paying attention to. And so I've started working on a very different paradigm, which is the Fitbit paradigm. So I would like to be a smaller, less substantial person. For a very, very long time I have been wearing this thing in my pocket, and it helps me figure out whether I want to change my behavior on any given day, get a little bit more walking in. And what I'm finding is that I now want a Fitbit for media. I want to find some way to let people see what they're paying attention to and what they're not paying attention to. And that's what underlies most of this research around the media cloud products, is trying to figure out, could we make it possible for you to monitor what you're likely to be knowledgeable about and not knowledgeable about based on the media that you end up consuming? And my students have started picking this up in really interesting ways. One of my doctoral students, Nathan Matias, makes me show this slide using my own data, because it's a really nice way to put me on the spot. He's written very, very good open source tools called Open Gender Tracker that make educated guesses at the gender of someone that you're reading. And looking at who I follow on Twitter, I follow a lot more men than women. And this turned out to be the case for almost everybody who has ended up using this. I've actually found many of the women in my professional circle will check two accounts. They'll have a professional account, where they're following twice as many men as women. They'll have a personal account, which is much closer to 50/50 or further along. But the whole idea is when you can make this visible, you can start making conscious choices on this. I am a lot less likely to follow some random dude on Twitter than I am to follow an interesting moment who comes into my stream because I have to show this slide every time I talk about this, and I'm embarrassed by it every single time I do it. I have another student who's been doing wonderful work trying to figure out whether we can use the web and some element of self-monitoring to try to bring some serendipity into what you encounter online. And so she's built a tool that looks at what news you read, figures out what parts of the world you're reading about, and then when you open a new browser tab simply sends you to a different city, gives you the city map, gives you reading suggestions, tries to push you in a direction of reading about something unfamiliar. What I'm feeling really good about in this is that I'm not alone on this one. Of the things that's been very interesting is to see Sir Tim Berners-Lee, 25 years into the web, standing up and saying, I'm not sure we got the web we want. I'm not sure that the web that we got, in Sir Tim's case, is a global enough, is open enough, is free enough. And he's now trying to get people together to look at this question-- do we have to be stuck with the web that we have? And I think this mission, this vision of essentially saying, we could fix what we have, we could make something very different-- if you share my sentiment that it's important to have a more global view of the world coming out of it, people in this room are actually in a remarkably good place to think about how you would actually go ahead and rewire and try to think about how you would put together a web that was better at making us globally knowledgeable and more cosmopolitan. So that's what I'm working on. That's what I'm thinking about in this book. The book is now out in paperback. So if you had been restricted by thinking about whether my ideas were interesting enough to carry 1 and 1/2 pounds of them around, it's now down to about half a pound. So if I've hit the half-pound threshold, you should definitely go get a copy of it. But I really appreciate you listening. I'm happy to take some questions if anyone has any. [APPLAUSE] Please. AUDIENCE: The first thing is the guy at Raytheon who found that the bridge people are more creative-- has anyone replicated that looking for the direction of causality? Because my suspicion is that many of the people who gravitate to those roles have a certain kind of way of thinking. ETHAN ZUCKERMAN: Yup, yup. AUDIENCE: And then the other one is, you seem to focus a lot on geographic diversity. ETHAN ZUCKERMAN: Yeah, absolutely. AUDIENCE: I mean, is that just a limit you have because of how you think about the problem? And do you know anyone else who's thinking of it as non-geographic in terms of diversity. I mean, you did the gender on Twitter. ETHAN ZUCKERMAN: Yeah, yeah. yeah. Yeah. No, thank you. That's an awesome pair of questions. So the Rob Burt work has been looked at very, very closely. That's a paper with like 5,000 cites to it. What most people have looked at is actually Raytheon's total inability to use the suggestions he made. So essentially having identified the smart people in the company, why has Raytheon continued to be such a disaster? I don't know specifically of someone who's verified your hypotheses, but I think it's really nice, which is, you may have certain people gravitating to that place. In the book, I use Burt to talk about this idea of a bridge figure. And for me, a bridge figure is someone who's sort of straddling two different worlds. And it's not just that you're bicultural. There are a lot of people who are bicultural who choose not to be a bridge. They're not actually interested in brokering between the two. I think there probably is a bridge personality. I wouldn't be surprised at all to find that bridge personalities gravitate there. You're totally right on the extent to which I'm using geography as an imperfect proxy for diversity. And what I end up talking about in the book is I go further with it, is essentially saying, what I'm really interested in is cognitive diversity. I'm really interested in this question of, do you look at the world in a different way? By spending a pretty substantial chunk of my adult life in West Africa, I can definitely tell you that the average West African has a really different and sometimes more hopeful set of cognitive tools to take on different problems. And it has to do with an educational system. And it has to do with the society that you live in. But there's no guarantee that that geographic difference is going to turn into a cognitive difference. You can imagine taking a friend of mine out of a West African primary school, educating them in the same schools that I've been in, working the same jobs that I do, we're going to end up thinking a lot alike in many ways. There are definitely people looking at questions around gender diversity, around racial and ethnic diversity. Over in our lab, the student of mine, Nathan Matias, has been doing really good, deep work on gender diversity and representation in media. And there's actually a very helpful community around there trying to get a sense for why is it that it's only 16% of op eds are written by women? And is this a place where you could sort of put a lever and try to get not only gender diversity but cognitive diversity in the process? Other questions? AUDIENCE: Yeah, I have a question about-- I like your comments about connectedness in the world, right? And to some degree, I think you expressed maybe some disappointment about how even the connections that there are internationally, that they are sort of obvious to some degree, right? Me as a German, I might be more connected to other Germans, so these connections are worth less to some degree, right? I wonder if we have enough tools for these kind of things. Like I mean, you mentioned like the cafeteria model, right? You go into a cafeteria and you sit next to people that are similar to yourself, right? And I think in social networks, it always tends to be that all the new connection you make are obvious given the kind of things you do, right? You travel to a remote office and then maybe you get connected to people in that remote office, right? Do you know if there's any good tools to get randomly connected, like sort of like a sort of virtual cocktail hours, where you just socialize with completely random people that also are interested in socializing with other completely random people? ETHAN ZUCKERMAN: So there's two approaches to this problem. There are people who've done pure randomness. And pure randomness doesn't work necessarily all that well. To a certain extent, Chatroulette is pure randomness, right? That was kind of the theory behind it. Let's just open a chat window, someone else arbitrarily in the room. The emergence of the social norms of, and now let me show you my genitals, was sort of emergent on top of that platform. But that may simply be where you end up when you're dealing with sort of pure randomness and pure anonymity. And there are people playing with it. And then there are services like StumbleUpon that are sort of trying almost pure random. There's an app called Random that works on this. The ones that I'm actually much more interested in are trying to engineer serendipity. And they're looking for a connection where some aspect of it is something in common, and some aspect of it is something unexpected. So two really quick models on this. Like a lot of people, like a lot of geeks my age, I miss Usenet. And I miss sort of the early '90s Usenet. And what I specifically miss about the early '90s Usenet was that it was a topic-organized internet. So if I found myself on rec.photo, there was a good chance that I was going to run into someone from Japan or someone from Finland, but we were going to have enough in common round emergent digital photography that we would have a reason to have a conversation. And then we might have more of that geographic diversity, maybe more of the gender diversity, more of the income diversity coming from having that topic in common. Where I'm now sort of thinking about this is I actually think Twitter is potentially a very interesting tool for this. Because the connection is pretty lightweight. You're sort of sharing information. And so part of what I'm trying to do-- one of the things we're playing around with right now is, can I try to figure out what your community of interests looks like on Twitter? And can I find a community that is related but sufficiently different? So if I'm a social media guru or whatever the hell that means in the US, can I find that person in Nigeria? And in having that element of difference as well as that element of similarity, can there be enough that we're likely to have that really profitable connection, but enough difference that we're likely to have something else of it? It's a really hard problem. And in doing sort of reading recommendation around it, which is what Catherine D'Ignazio and I spent a lot of time doing on that Terra Incognita problem. We actually figured out that it's probably not as simple as hi, have a chat. It's probably more like, can I give you enough information that you would have something to have a chat about? And so the model that I'm in some ways really interested in is one that started in Scandinavia, is now sort of playing out in public libraries, which is the human library, which is you identify a set of people from very different walks of life, very different backgrounds. You put them in the library, and you offer someone the opportunity to go check out a person and sort of have a conversation. And whenever this has been done at libraries, it sort of sells out almost immediately. My wife and I were sort of arguing over who got to do it at Williams College in Williamstown, and we figured that random professorial type was not as appealing as rabbi, which was her identity. And so she signed up as a rabbi and had like 9 of her 10 slots filled almost immediately. The one guy who was more popular out in deep blue Western Massachusetts was Gun Owner. And everyone wanted to go talk to this guy. And I'm now actually taking shooting lessons from the guy. So there are really interesting things that sort of happen when you set up structures to allow people to meet. But I agree with you. I think in many cases, it's something where we would want to do a lot more work. Please. AUDIENCE: This is sort of a modeled question, but I think-- well, my viewpoint anyway-- we have a bit of a problem with this partisan echo chamber. You can select news sources that are never going to deliver you information that challenges your worldview. But if I go look for information sources that are not from my partisan perspective, I wind up reading op eds that sort of make me want to throw a brick through my computer screen. And I don't know-- do you have any thoughts on if there's a way out of this trap of you want information that challenges you, but you don't want information or opinion, maybe, that is repugnant. ETHAN ZUCKERMAN: I think that's exactly right, and by the way, not a muddled question, a nicely put question and right at the heart of some really interesting debates going on in the space. There's a huge debate going on about whether the internet is separating us politically. And I would say on one side, probably the best articulated is a guy named Cass Sunstein, who has said, I've done experiments in deliberation. You put people in the room with like-minded people, people get more extreme. If I hang out with the far left, I get further left. You hang out with the far right, you get further right. Gentzkow and Shapiro, really pair of economists at Chicago, wrote a paper essentially saying, no, no, no, the internet's actually much better than the rest of the media. You're actually going to get more diverse audiences. It's a really controversial paper. I take it on at some length in the book and come to a different conclusion than they come to. But they're better economists than I am, so they may be right. I may be wrong. My take on this would be the first instinct, which is, OK, let me turn on Fox News, and I'm going to understand what they're thinking about, is probably wrong. Because you're going to the absolute opposite pole of where you are. And it's really hard to look at this other than, that's the enemy, I want to figure out who they are, and I want to undercut them. I actually think the thing to do is, when you're sort of orthogonal with someone, to try to figure out how you get the optimum sort of angle of consonance, right? So if I'm looking for a conservative viewpoint that I might want to pay real attention to, I'm not going to look for someone who's contemptuous of science and someone who's deeply religious and someone who's from the American South. What I might end up looking for is someone who is also from New England, also in academia, but happens to think really different things than I do. And when I'm able to find those people, I find myself much more challenged. And so that may only be a 15 degrees of difference rather than 180 degrees of difference, but that 15 degrees of difference puts me in a place where I'm actually forced to deal with the argument and sort of read it through. And so what I'm interested in is, can we start identifying those people 15 to 30 degrees off of where we are, and then find out what they read, and find out how we meet them, and sort of find out how we sort of have those conversations? These are the sorts of systems that I would really like to be building. But I've only got sort of three or four grad students over there. And this is a big company. You guys build a lot of different things. I'm really interested in, can I get other people in the tech industry who are interested in exactly that question that you're asking to sort of say, is this perhaps part of our responsibility? Is this part of what we should be asking our tools to do, is not just give us what we want, but possibly giving us what we need in civic terms to try to get to a more inclusive society in one fashion or another? Anybody else? Hey. AUDIENCE: Hello. So people listening to this talk who may be very interested, but they're working on another project, and a few months pass, and they forget about it-- how can we foster a more frequent contact so that this stays on people's minds? ETHAN ZUCKERMAN: Well, I am across the street. And you can come see us. And you can actually get into to my building without ID, which is always nice. But so a couple things-- I sort of soft-pedaled the Global Voices project. But there's actually a very vibrant community of people around that really working on this question of, how would you pay more attention to the world and sort of how would you fight the way in which those countries are framed and understood? And so one of the things that I would definitely encourage people to do is to think about, is that something to connect with? We are doing some active experiments on trying to figure out both, can we monitor media consumption voluntarily, sort of opt in? And when we learn that media consumption has certain biases associated with it, can we do structured recommendations of content? And for me, that strikes me as a very interesting place-- potential to collaborate. And we do have students sort of working on that and sort of playing with that, and you guys would be very welcome to come across the street to the Media Lab, have conversations about that and sort of figure out where to go. The trick in sort of writing a book is you're trying to identify people who think a problem is important enough that they want to put some time into it. And so this is sort of my plea to people who think this is interesting enough that they want to work on it. And I'm now sort of advising half a dozen projects around the world of people who think this is an interesting thing to explore and want to work together on it. I would love to be involved with sort of more of those conversations. I would also love to be involved in conversations with people who I've got it wrong and I'm not framing the question the right way or the most provocative way. But I would hope to sort of open a door around this, either to that existing, thriving community around Global Voices or to the nascent community around civic media that we have across the street. Please. AUDIENCE: I think also that might be a good opportunity for the plug. If anybody here is interested in sort of topics like citizen journalism and stuff like that, we have a team here in Cambridge working on YouTube live streaming, where we sort of worry a lot about news and how live streaming of news sort of can impact these kind of questions around news and citizen journalism and stuff like that. And we're now starting work with MIT Media Lab, actually. YouTube is funding research at the MIT Media Lab to sort of explore these ideas. So we've been actually in contact with Ethan, people who are interested to work on these things. ETHAN ZUCKERMAN: And precisely the project we're talking about working on together is trying to figure out, how do you put enough context around the live stream of an event that it's comprehensible to someone who doesn't know where it's coming from? I was looking at this the other day, and the protests about the internet tax in Hungary were being live-streamed on YouTube. And on the one hand, this was sort of fascinating. It's really interesting to see tens of thousands of Hungarians come out into the street over an internet access issue. The flip side is that my Hungarian is not very good, and there's no information or sort of context going on on that. How do we turn that amazing ability we have to now be in the streets of Budapest watching that protest unfold? How does that turn into that opportunity to actually get informed on it? So we're excited to work with you guys on that. Thank you so much, guys. Please help yourself to a book. I'm happy to sign one if anyone is interested. I'm happy to sign more than one if anyone is interested. It's been really great to be here. [APPLAUSE]