 Good afternoon, and it's a great pleasure to be with you today, and what I'm going to be talking about is the future of broadband. You heard earlier from Blair Levin, who's the man from Washington, D.C., who really understands policy. I'm the guy from Silicon Valley, and I'm going to talk about our view of the world, which is somewhat different. I work at a small organization in Palo Alto, California called the Institute for the Future. We've been around now, I think, 46 years. We're founded in 1968, doing long-range planning and forecasting, and in fact have worked with a number of the people in this room over the years. Our motto is foresight that leads to insight that leads to action, and we see ourselves as in the business of providing the foresight, and we work with our clients to generate the insights, which hopefully will lead to the action, which is their responsibility. As I was preparing this talk, I was thinking about one of the four founders of the Institute. There's a man named Paul Barron who passed away in 2011. He probably has as good a claim as anybody to truly being the person who invented the Internet, probably more than Vice President Al Gore. In 1964, Paul published a paper for the RAND Corporation, which is where he was working, called On Distributed Networks. He began by describing three kinds of network structures, what he called centralized structures, decentralized, and distributed networks. The reason for this is it was actually commissioned by the Air Force who was interested in what kind of a network architecture would survive an enemy attack. A centralized network was vulnerable because it could be taken out from a single point, but a distributed network had no center and therefore was much more robust, but it turned out that that really provided the kind of DNA that really underlies the fundamental architecture of the Internet. And as I'm going to argue, we're going to see this diagram come up later on, is becoming more and more the blueprint for society in the larger sense. So what I'm going to talk about in the few minutes that I have is, first of all, I want to talk about the power of exponential change. And I think that that's a driver that is sort of a challenge for us all to grasp. Then talk a little bit about the evolution of the Internet, the future of broadband, and then go a little bit beyond broadband to say what it all might mean. And then end up with a few suggestions that I hope will be useful. Well, let me start with the question of exponential change. Human beings in general can understand linear change or incremental change. We live with that, but we really have a hard time with the concept of exponential change. Although I think the Qataris probably have an advantage here because they're living in a world of, I think, of exponential change at the moment. But where this idea really came from was Gordon Moore, one of the founders of Intel, who back in 1964 formulated Moore's law, which is really not a law of nature but an observation. And what he observed is that the number of transistors per square inch on an integrated circuit had doubled every year since the integrated circuit had been invented, and he predicted that this would continue to happen in the future, so doubling every year. And in fact, that prediction has turned out to be remarkably true. So if you just look at the math of that, take a number and double it every year, what you get is something that looks like this. And those are years at the bottom. It's 1964 there on the left when he made that observation. And what's interesting is for a long time that line goes along fairly level and then it starts turning up and it goes nearly vertical. And that's where we are now. We're on that upturn of the hockey stick. So we're at the point where at least in terms of the fundamental technologies that are driving us, they're all at that point of these very large numbers. And so let's talk a little bit about what that means. So in terms of raw computing cost, what Moore's law says is that the cost of digital technology is going to keep falling. And in fact, over a 20-year period from 1992 to 2012, the cost of raw computing power fell from $222 per million transistors to 6 cents, which is an improvement of 3,700%. And clearly this enables computing power that's at the very core of the digital infrastructure to get far more powerful. And it also gets tinier and tinier. So these are the things that we're carrying around, I assume probably everybody in this room, carrying around these devices in our pocket, have these things that are beginning to approach the power of supercomputers. Same is true of storage. Over that same 20-year period, the cost of data storage has actually fallen even more sharply from $569 for a gigabyte of data to 3 cents in 2012. And that's a drop of 19,000%. So storage is getting to be almost free. And finally, bandwidth cost performance. The numbers aren't quite as dramatic, but they're in the same direction. Cost fell from $1,200 roughly for 1,000 megabytes per second in 1999 to $23. And this is really what is speeding up the flow of information. So all these are big numbers. Let me give you a concrete example of what this means in practice. And we'll go back to 1956. And that device that's sitting there on that forklift is the world's first hard disk drive. It's the Ramac 350. It was built by IBM to accompany one of its systems. It had something like 24 separate disks in it. It had a total capacity of 5 megabytes. And every time I take a picture with my mobile, it's more than that. The engineers and IBM, by the way, said they could make a hard drive with 10 megabytes capacity, but the salesman said there was no market for it, so 5 megabytes was plenty. And at least for about $3,000 per month. So let's fast forward to 2013. And on the right is something called the Oyan Shadow Drive that you can carry around in your pocket. It's 5.2 ounces, which is one 6,000th of the weight of that Ramac. And if it held even the same amount of data, that'd be kind of amazing, but it has a capacity of a terabyte, which means that it holds 200,000 times more data than that Ramac did. And finally, the Oyan Drive, you can go on Amazon.com right now and buy it for $89.95, unless it's on sale now. It might be less. So the cost of storage on one of these drives is about 6 cents per gigabyte, or a tiny fraction of 1 cent per megabyte. So storage almost free. I'll give you one other quick, one more example. In 1994, Apple Computer, when they're often trying to push the envelope, they released one of the world's first digital cameras. It was called the Quick Take. It had a resolution of 0.3 megapixels. It could store a total of 8 photos. It had no focus, no zoom. There was no way to preview any of the pictures. You had to download it to your computer and then look at them. And I managed to borrow one of these. I was a fairly serious photographer and I was interested in this idea of a digital camera. So I borrowed one, went around and took some pictures and I downloaded it to my computer. And they were terrible. They really sucked to use the technical term. And I went around predicting to all of my friends that digital photography will never replace film. Wrong. I don't think there's any film cameras here today. So here we are in 2013. This is the Sony RX100. It has 20 million pixels. It has a 28 to 100 millimeter zoom lens. It has an LCD screen with a million bytes in it. An ISO rating that goes up to 6400. And it also does video. It does panoramas. And here it is right here. And of course this is an obsolete technology also. This is because nobody really wants a separate camera. It's disappearing inside of our mobile phones and not only will they be soon better than this camera, but it's connected to the world. So I've already, like this morning, sent this audience and sent it back to my kids in California so they're following this event in real time. So that's an example of the way in which technology, in a very short period of time, can start doing things that a few years earlier had seemed impossible. So now let's talk about... Oh, those are the examples of the pictures that you can take on the left. That's what the apple took on the right. And that's actually a zoom, just a small part of one of those pictures. And I promise that's the last picture but I wanted to bring them here. So now let's talk about the evolution of the Internet. So here we are in 1969. This is the full extent of the Internet. It's the first deployment of what then actually was known as ARPANET. There were four nodes, one at Los Angeles, one at the Augmentation Research Center, at Stanford Research Institute, headed by Doug Engelbart, one of the sort of secret heroes of the Information Age, one in Santa Barbara and one at the University of Utah. They were all interconnected with modems on least lines that ran at the breathtaking speed of 50 kilobits per second. And there were, I don't know, 100 people who were on the Internet. And here we are today. I mean, I can't show a map that shows all of the connections, but this is a kind of a heat map. This is actually going over a 24-hour period and the redder the colors are, the more intense the use is. It's a daily pattern. And you can see that wherever there are people and electricity, there is now Internet activity. And certainly the northern tier of the world is extremely hot. India and China are coming along. Africa is still in development, but that's where the next billion or two billion people are likely to come from. And this went from this tiny little thing to taking over the whole world. But another example of how rapidly this world is changing, and it's not just people, but it's the things that the Internet can do. And just go back, just think about going back to January of 2006. So at that point, Facebook was only on college campuses. There was no, none of us had access to it unless we were a student in college. Twitter didn't exist yet. It was created in March of 2006, and it was launched in July of that year. So nobody had ever heard of Twitter, which is running over there. In 2006, Netflix was the largest single customer for first-class mail delivery of the U.S. post office. In that year, they had mailed their billionth DVD. By 2010, they had become the biggest source of Internet traffic in North America in the evening as they moved to streaming service. So that entire business model had shifted. There was no iPhone in 2006. Nobody had ever seen that kind of device before. It was released in 2007, and Android didn't come along until 2008. And so the smartphone was just really not something, you know, the Blackberry was about as far as the technology went. Nobody had ever heard of a MOOC in 2006. Coursera, which was the thing that grew out of the Stanford course that had 160,000 people online, was launched in 2012, and it was been joined quickly by many others. And finally, in 2006, clouds were just something you saw in the sky. Amazon Web Services was introduced at the very end of 2006. It now serves over 100,000 companies in 190 countries, and really helped to spark the whole movement of cloud. And I think I've seen data that says now half of the world's computing is now taking place in the cloud. So okay, that's the story of change. So now let's talk a little bit about the future and what it might hold. And given my scenario, what I'm saying is it's really very hard to know what the future will hold. It's going to be full of surprises, but there are some places, as Blair said, there are some places where the future has already arrived. We can look at those. And one of them is in the scientific research world. The National Lambda Rail in the U.S. is an ultra-high-performance network that supports advanced research activities at hundreds of universities and research laboratories across the U.S. It's currently offering connections at 40 gigabits per second, and they're working on moving up to 100 gigabits per second shortly. And what drives it is what they call application-empowered, high-performance e-science projects, high-energy physics, astronomy, earth science, bioinformatics, and environmental sciences. And what they believe is that in the coming decade, e-science is going to require terabit networks, exabyte storage to support distributed peda-ops computing. So, you know, these notions of terabits and exabytes and peda-ops, it's like a whole new language. You know, I think we've already learned about, you know, we've moved from megabytes to gigabytes to terabytes, but now, you know, these next increments, and every one is an increment of a thousand. And I think Robert Pepper probably will talk about some of these numbers and what they mean this afternoon. So, what's going on there? One of the most spectacular uses of the Lambda Rail is a multi-gigabit network known as SAGE, the Scalable Adaptive Graphic Environment, which makes it possible to create wall-size displays of scientific data by tiling arrays of large video monitors that permit users to visualize and interact directly in real time with data in many novel ways. And this is currently being used for scientific research, but I believe that versions of these versatile displays will eventually be found in businesses and homes. And Dr. Pepper told me, oh, if I come down to visit Cisco and Santa Clara, he can show me a commercial prototype that they've already got running, which is a great example of the future usually arrives faster than we expect that it will. Another interesting use of Lambda Rail is CineGrid, which provides ultra-high-speed connections that support digital media exchanges among remote participants. So, for example, you could have a worldwide crew that are editing a movie together in real time with zero lag, an example of what Blair would call high-performance collaboration that can be done seamlessly over the network. So that's the national Lambda Rail in the U.S., but there's a bigger picture, which is the global Lambda integrated facility, GLIF, which spans the world, and it provides a network that makes it possible for scientists around the world to get rapid access to vast amounts of data, for example, from the Large Hadron Collider, which sits in Switzerland and generates enormous amounts of data that can be transported over something like this network. The participants in the GLIF are the National Research and Education Network's N-Rens, and I noticed that part of the broadband plan for Qatar calls for the establishment of a national research and education network in Qatar. And so that means that you will be connected to this global facility. Another development that we can see that's going to change the nature of networking in the near future is what's known as software-defined networks. OpenFlow is a standard for software-defined networks that allows the path of packets through the network to be determined not by hardware, but by software that runs on multiple routers. And so what this means is that we're going to move from a static, essentially one-size-fits-all internet to a more flexible architecture that's controllable by the users themselves to optimize advanced application. And this is going to increase flexibility tremendously, but it also can raise the specter of network fragmentation and even vulcanization if everybody is running their own network. This, by the way, is the architecture that sits inside many of the largest data centers. Internal data centers are entirely run on this kind of architecture, but up till now the public-facing networks are different. Let's go a couple of more examples of the things that are appearing and that I think suggest where the future is going. Innocentive has been described as the eBay for innovation. It's an open platform where companies who have problems that they're struggling to survive can post what they call challenges and invite anybody in the world to offer a solution in exchange for cash rewards. And they've had thousands of problems posted and thousands of problems solved and the basic premise is no matter how many smart people you may have working for your organization or your company, they're more smart people in the world at large. And Innocentive gives you access to them all. Another example that I really like, a brand new company that's in Palo Alto, California, is Declara, which is a kind of a smart social network for organizations that uses machine learning to understand what the capability of the participants are of the network and then matching them up together. The first deployment is among the 280,000 K-12 teachers in Australia that are giving them a platform to share best practices and ideas. And so instead of each teacher in a classroom kind of working on their own, now they'll have access to all of the best experts in their country. And there's no reason why something like that won't go global. And finally, two weeks ago, IBM announced that Watson, which is the system that they created that went on the quiz show Jeopardy and beat the two human challengers a couple of years ago, kind of a very elaborate stunt, but they've now broadened its capabilities and they announced two weeks ago it's going to be online and anybody can sign up to use this. So instead of thinking of a computer, something that you program essentially at a low level or a high level, you'll be able to go in, you'll ask questions and Watson will answer it. And the people there are suggesting that this is changing the paradigm with which we're going to interact with computers. So again, models for change that really each one can radically change, but that's not all. There's a few more. Google Glass. There are already lots of people walking around my area out in Silicon Valley who are wearing these things that are, and I've tried it on. It's a little bit strange, but you're bringing the internet with you. You talk to it or you touch it briefly. It's on your eyeglasses and it'll give you the answer to any question. And again, this is just generation point zero. You think about like 2022, the people who are going to be coming to the World Cup, many of them are going to be wearing the great-great grandson of this device. So it's not just going to be a better phone in their pocket. It's going to be incredibly powerful devices that they're going to be wearing. And what their expectations are for the kind of connectivity they want may be sort of a surprise. One of the things that you're going to be doing is augmented reality. You'll be able to overlay the real world with data. And in fact, one of my colleagues at Institute for the Future says, we'll look back on this era where we walked around in ordinary reality as sort of quaint and outmoded in a few years. We'll just simply expect the data to be there. And I think it's very interesting that Qatar is essentially scanning in 3D the entire country and will have the geo-coded data above ground and below ground that will give you the ability to do this kind of virtual reality very easily. Driverless cars are coming. I was in a parking lot in the little shopping center next to my house a few weeks ago. And there was a guy from Google taking a smoking break outside his Google car. And he just spends the whole day kind of running around on the freeways, getting experience. And I asked him, I said, did he keep his hands on the wheel at all? And he said, near them, near the wheel. And I asked him, was he ever scared? And he said, only by other drivers. So it's working quite well. And I think within a decade that's going to be real. Robots are starting to arrive. This is a little device that sells for under $200 called Autumn. It's a little assistant that provides support for people who are dieting for weight loss. So it's very friendly, very simple, but of course the power comes from its connection. And the news just came out last week that Google is making a very, very large investment in robots that I think is going to be an interesting watch. They haven't said really much about what they're doing other than they're doing it. One week ago, Jeff Bezos, the CEO of Amazon, was on television in the States. And he announced that one of the things that Amazon is playing with are pilotless drones because he wants to be able to fulfill an order within 30 minutes. So you'll order something and within a few minutes this drone will arrive in your backyard and deliver the package. And finally, as I arrived in the airport here, I saw this poster from HSBC that has this campaign about the future. And it says, in the future, exports will not be transported, they'll be transmitted. And at first I had no idea what that meant, but it sounded pretty futuristic. And then Dr. Pepper told me, well, what they're talking about is 3D printing. In other words, if you want to get a part delivered, you don't have to send the physical part, you just will send the blueprint for the part and a local 3D printer will print it out. And what's interesting about all of these applications is this is taking the Internet and now moving it out into the world in kind of autonomous form. So one of the things I think is really happening now is that the Internet is rapidly transforming from a place, a sort of a separate place, cyberspace, that we would go visit for specific purposes to something that's just simply part of our everyday environment, that's going to be in objects, that's going to be projected into the world. So it's going to be a constant companion for us. So now let's move beyond broadband. So given all of this marvelous technology, what does it mean for society? And this is really, I think, the most critical element. I would argue that we're now in the process of moving from a connected world, which is what we're talking about here, to what I would call a hyper-connected world. And what are the implications of that? Well, one of them is my colleague from the Institute for the Future, Marina Gorbis, has written a book called The Nature of the Future when she talks about what she calls a social-structed world. She means that individuals or small groups are able to harness the power of technology to join together in new ways and provide an array of new products and new services in ways that were simply impossible. I think the Khan Academy is a great example of that one man with a video camera and a pen who was trying to teach some math problems to his niece in a few states away ended up creating a global resource now that's changing the nature of education everywhere. And these new kinds of enterprises, again, what we're seeing is these diagrams, which we saw 50 years ago from Paul Barron, now really representing the way in which individuals relate and in which structures are created. Another example, another group that's studying this new world that's being created by broadband internet is called the Deloitte Center for the Edge, again located in Silicon Valley. It's headed by John Hegel and John Seeley Brown, the former director of Xerox PARC, and one of their most recent studies looks at how exponential change in technology leads to what they call exponential innovation. They also have talked about something they call the big shift. And I'm going to show you a picture and if you walk away with anything from my talk, I think this picture is probably the most important. And it's pretty simple. It's the shift from stocks to flows of knowledge. So if we think about the 20th century, this was the knowledge century where we would accumulate and we would build up stocks of knowledge and then we would be able to exploit them. Think of patent portfolios. Think about companies with great intellectual property, companies that have reached economies of scale and have been able to turn out and mass produce goods to the world, but in a world of constant and accelerating change, those stocks of knowledge decay so rapidly that you can't exploit them over time. In certain fields, technical fields in school, students who are in college for four years by the time they graduate, half of what they learn may be obsolete. And so there's no choice but to engage in a world of continued learning. So we have to move, they would argue, from a world of scalable efficiencies, the path to success in the 20th century, to pursue scalable learning, which enables everyone to become better, faster by actively participating in networks, both internal and external, and have opportunities to take on and deal with really challenging problems. So that innovation, which we think about in the world of products and services, is great, but we also need innovation in our institutions to evolve to work in this new world. Now I'll just give you a couple of quick examples of what I mean by this transition from stocks to flows. Consider Encyclopedia Britannica. It's great work of accumulation of man's knowledge, first published in 1768, published continuously until 2012 when the company that published it announced that was the last print edition of the Encyclopedia Britannica. And why? Well, it was something called Wikipedia that came along, crowdsourced online Encyclopedia that turns out that people kind of look down on until they realized it's much larger, it's much more up to date, and even more importantly, totally transparent because on every page you not only get the information but there's a tab up on the top called Talk. And if you click on that tab, you see all of the discussions that go into the making of that entry so that you're not only given the information, you're given the story that's behind it. Another example of moving from stocks to flows. Education used to be something that happened inside a school, and that happens to be the high school in Denver, Colorado where I got my high school education and it was a great place. But, you know, today, education no longer is confined to the walls of a school. We are now creating what John Sealy Brown calls Learning Networks with an infrastructure that can unite not only the schools but museums, after-school programs, homes, and online resources, and the schools just really become one more node in a learner's learning network. And it's really the learner, him or herself, who can be at the center of their own education and pursue whatever interests them and takes them on. One last example, you know, it's a great institution of our newspapers which package the world up and would deliver to us every 24 hours. Now it's seeming increasingly anachronistic. You know, we have 24-hour news, CNN, Al Jazeera, but, you know, increasingly we have Twitter feeds so that you don't need the newspaper at all. We have direct reports from the world from the people who are participating in it. So we're, you know, moving, again, from stocks of knowledge to flows. And so the issue is really one of participation. So that's the opportunity. Here's the challenge, though. You know, technological change is exponential and it's happening very fast. But social change, business change, and particularly political change happen more slowly. And this is, I think, this phenomenon accounts for the reason that Blair explained that the most important sentence in the national broadband plan that he created is that the plan is in beta and always will be because it's going to have to keep changing. So let me end up with a few suggestions I hope will be useful to you. A couple of things that I think it would be good to do, some things maybe not to do. One of the things not to do is what the U.S. just tried to do with healthcare.gov. I don't know if this disaster has made its way around the world, but it's been an enormous embarrassment for millions of people. And I happen to think that the product they're offering is absolutely incredibly important and valuable. I mean, to extend healthcare coverage to millions of people, U.S. being one of the few countries that doesn't provide this for everybody, but the implementation, as Blair said, got messed up royally. And, you know, an example of trying to use very old 20th century, if not 19th century processes to build something that was a disaster. You know, a better example, one that I really like is something called the Direct Project that was also done by the federal government. For over a decade, the government has been pursuing something called the National Health Information Network, NHIN, which is a vast, very elaborate, very ambitious infrastructure to connect all the healthcare systems. And they progressed very slowly. And the only reason you haven't heard about it is because there was no mandate for it to ever have a deadline. But in 2010, a small group of entrepreneurs, both inside and outside of government, decided they were going to try a different approach. Rather than trying to develop a set of specifications, issue a contract, and then run the contract, what they did was to set some very high level goals, set up a public wiki, and then invite anybody who wanted to participate to come in and help them design the product. And within less than a year, they had attracted, first of all, about 65 major players in the U.S. to take part. Within less than a year, the project moved from a discussion of goals to creation and testing of a prototype to its initial rollout, and now it's being taken up as very limited, modest goals. But it's been a tremendous success. And it was an example of how government can function entrepreneurially. Another example, I think, of what to do is in Mexico City earlier this year, they set up something called the Laboratorio para la Ciudad, or the Laboratory for the City. It's an experimental program whose goal is to foster civic innovation and urban creativity. For example, they partnered with Code for America, which is a program to recruit young people to help governments to create Code for Mexico City. And they asked for proposals to solve city problems. They got applications from some 250 young people who wanted to work for city government, and now they've put them in a room and they're working on all kinds of interesting applications. Another example of a good thing to do, Google Fiber a couple of years ago went to Kansas City and said they were going to wire up the whole city with gigabit speed fiber, which is great. And then the city responded by saying that this was a wonderful gift, but that they needed to do strategic thinking among the citizens to figure out how to make use of it. Another great example of what to do. This is one of my favorite places in New York City. It's a storefront in the Chelsea neighborhood of Manhattan called Senior Planet, and it's a showcase for cutting-edge technology for senior citizens. And so seniors are invited in to play with the technology. They can take workshops and classes in the back room, and every time they have a sign up for classes there's a line that goes out the door. And a final example, if you're trying to cultivate innovation in Qatar, it's very appealing to think of it like agriculture. You prepare the soil, you plant the seeds, you till the ground, you fertilize it, and the wonderful crops will come up. But that's really not how innovation happens. It doesn't really work that way. There's really another model, and it looks more like this. It's two young entrepreneurs in Silicon Valley have argued that it's really like a rainforest, not a farm. It's disorderly, it's teeming with all sorts of life. It's very hard to tell the weeds from the seeds. But, you know, when Facebook started it didn't look any more promising than any other of a thousand startups which didn't go anywhere. So, you know, if you want to really cultivate innovation you have to do things like create a culture that will support it. There's a long conversation about what that involves. But that's a discussion for another time. Let me just end with a final thought. So two very different views of the future. And the first view comes from that great American philosopher Garfield the Cat, who is the hero of a comic strip. And Garfield's a rather sedentary, pleasure-loving character. But he has a lot of interesting thoughts. So in this strip his owner, named John, is looking at Garfield, who's in a characteristic pose. And he's saying to Garfield, you know, we can't do anything about the past. But then he says, we can do something about the future. And Garfield lies there and he thinks, it sounds like a lot of effort to me. He says, I think I'm going to like the future just the way it's going to be. So, you know, if that's your view, life is simple. On the other hand, there's another philosophy that I like better that has been attributed to Peter Drucker. And that is the best way to predict the future is to invent it. And I think, you know, Qatar has a wonderful opportunity now to invent its future. And so I want to wish you fair winds and calm seas and a wonderful future as you move into it. Thank you.