 Well, let me open up by just talking about the two faces of technology. So on the one hand, what you have is an unprecedented time at which the American people have the ability to access their legislatures, to access the process of actually writing the code that governs your country. On the other hand, you've got a divide, whether it's the digital divide or actually the entire process by which legislation is created, language that no one can really understand, the average person doesn't really get. And also, in the context of platform, I'd like to talk a little bit about that. But in terms of the opportunity, as you saw President Obama during his election, our campaign process, he was able to raise online over half a billion dollars, with the average contribution being under $100. So he had the ability to mobilize an entire community. In the same way, we have a unique opportunity that hasn't really been leveraged, which is engaging the American people in the legislative process. In terms of thinking through how would you write pieces of legislation? How can you use these new platforms that have never existed before to really empower the American people? And how do we even rethink the process of actually writing legislation, having people improve it over time? Imagine a world where you've got almost like a Wikipedia, as you think about some of these really, really complex issues. And as complexity increases, the way to solve it is to bring in the brain trust across the country to tackle on some of the most difficult problems. Yet I wish we had, when it comes to spending, a TCPIP protocol for how the United States government actually spends money. So a couple of issues that I've been dealing with directly, I remember sort of day one on the job, having to think about a common architecture of how we would track recovery spending across the country, a nationwide system. What we learned very quickly was there wasn't a common architecture for program codes. There wasn't really a common architecture at the federal, state, and local level. As a federal government, through the legislative process, begins to appropriate billions of dollars in funding. And we've done a lot more work in that space when it comes to usspending.gov. We just launched last month a new version, a new platform. And part of what we try to do is model that platform, very simple, very much like a Google or a Bing or a Yahoo, where you could literally go online, you could go online today and type the word wine, and you could see you'll get an image of a map, a national map that will show you where the United States government is spending money on wine. The challenge has been it's taken so much effort on the back end because to your point and to what Tim was talking about in terms of how this funding is appropriated. So there isn't a common architecture as we think about how the government spends. Also there isn't necessarily a common architecture when we think about performance. So after you move from policy formulation and where legislation passes to actually execution, where you're literally trying to implement programs, what we've noticed is a huge gap in terms of being able to shine light in those areas. So we've done that in a small part for the $76 plus billion around the IT budget, where we're literally putting online the picture of all the CIOs you're talking about next to the projects that they're responsible for and how those projects are performing in terms of cost, in terms of schedule, in terms of the stated outcomes up front when the agencies went in front of Congress to ask for that funding to really shine light so we can see and learn in terms of lessons as far as what works, what doesn't work, evidence-based approach to implementation. The other area in terms of platforms or recognizing is to Tim's earlier point, which is Apple created a platform the same way what we've done is we've created a platform called data.gov. And what data.gov does, it actually takes the vast array of storage or information that we have within the federal government without compromising national security or privacy of the American people, we're trying to democratize that information. We're finding some really, really innovative uses of government data. We've talked about before the revolution that was unleashed when the federal government decided to release GPS data where all of a sudden you have GPS devices on your iPhone and the ability to navigate a new city or town or going to your local car store and actually getting a GPS device. In the same way what we're seeing, simple innovations, for example, the Health and Human Services Department for years has had a site called Compare, which essentially allows you to compare the outcomes, those hospital compare, which allows you to see customer satisfaction in terms of patient satisfaction by hospital, outcomes by hospital. But it's one out of 24,000 websites in the U.S. government and had very little foot traffic. As soon as we democratize that data and made it available in machine-readable formats, what happened is companies like Google and Bing took that raw data. And Bing, for example, now if you go there and you search for Sibley Hospital or Shady Grove Hospital or any hospital in the country, right there you can see the patient outcomes of the hospital, how people have rated it, and that's really putting information, the fingertips of the American people so they can make intelligent choices. And we expect as legislation passes in different laws and different domains, imagine the ability to take that raw data or that information and think about the government almost like a platform in the context of a grocery store where anybody could take that data, mix the ingredients, and create all sorts of magic. So we saw that not just with the data we released at the Health and Human Services level, but also FAA data, we've got a community of developers at RPI who have built over 40 applications from applications that look at White House visitor logs and actually visualize who was visiting when and all the news stories associated during that time to taking data that has to do with the U.S. government budget and looking at what were the hot stories at the time the budget was being deliberated. So there's unprecedented opportunity when we look at the intersection of cheap compute power and the deployment of broadband, which for the first time has given the American people a front-row seat to their government. Imagine what happened in the Agora. It used to be that people would petition their government by convening at the town center, a public square, whether it was to socialize or to discuss issues a day. Now anybody anywhere in the country has the ability to petition their government to contribute, to challenge and to actually hold this accountable. And part of what we're trying to do is look at this in terms of the whole ecosystem from policy development to actually the legislative process when laws get passed. But most importantly, the implementation. How do we make sure that we're shining more light and releasing more and more information around the implementation part of the legislative process? Because once we know what works, we can continue to do more of that and divest from those things that don't work.