Microsoft’s Craig Mundie on Future Interfaces, Computer Science Education, and Life After Bill G
Craig Mundie is a geek, and I mean that in the best possible way. Microsoft’s chief research and strategy officer, the 17-year veteran of Redmond, WA, still talks like an engineer, throwing out terms like “heterogeneous machine architectures,” “GUIs” (graphical user interfaces), and “clouds and clients” like there’s no tomorrow. It’s kind of refreshing, given that he is in charge of setting the long-term agenda for one of the most powerful companies on the planet.
Mundie is in the midst of a weeklong tour of some top universities around the country. He called me yesterday from Cambridge, MA, where he had just finished a presentation to Harvard University students, faculty, and guests. He visits the University of Illinois at Urbana-Champaign (my alma mater) today, and comes to Kane Hall at the University of Washington tomorrow afternoon. It’s similar to the college tours Bill Gates used to do.
From what I can tell, the goal is to stir up interest in computer science, give audiences a glimpse of future computing systems as Microsoft sees them, and stimulate discussions about how these technologies can help solve some pressing global problems. (You can read more about Mundie’s tour and demos in this Seattle Times story.)
Besides hearing Mundie’s thoughts on computer science education and the future of computing, I wanted to drill down and ask him about the challenge of taking on Microsoft’s strategy development (after Gates stepped down last year) in the most difficult economic times in recent memory. I also wanted to ask him about the deeper culture of Microsoft, the renewed role of research in the company’s future, and the importance of nurturing relationships around the world—and his secret ally in that quest.
Here are some edited highlights from our conversation:
Xconomy: What are you trying to get across to university audiences on this tour?
Craig Mundie: In these presentations, I’m trying to get them to think not only about how computing evolves, but with that evolution, what kinds of problems will become approachable, and what are the new methods? Several things are evolving in parallel [and leading to more heterogeneous and complex machines]. That begets the requirement of how to do programming around parallel computing. With very high-scale computing facilities, the cloud and the client come together to form one system that people will program. They will use those things together with new display and sensing technologies.
Just as the GUI revolutionized computing, we could see a similar revolution with more natural interactions with machines, rather than just “type and point and click.” That will expand the number of people who can interact with computers. With the diversity, rooms can become computers [for instance]. You won’t think of them so much as a computer.
X: What are some of the global problems you think advanced computing will help solve?
CM: Beyond the computer science realm, I’ve talked about energy and the environment. I show one piece of research work we’re doing to compose computational models, a simplified climate model, at Princeton and Microsoft Research. It shows linkages between deforestation in the Amazon and atmospheric temperatures around the rest of the world. If you were a policy person, these kinds of things would give you tools to support your decision making.
In energy, we’re doing computer modeling and direct visualizations. I showed a model, loaned to us from TerraPower [the nuclear power firm spun off from Nathan Myhrvold’s Intellectual Ventures and backed by Bill Gates and others], and how they’re exploring the nature of nuclear.
I also showed a visualization with gestures as a way to couple myself into a computational model to adjust the parameters of a wind turbine blade. There’s no traditional GUI.
X: But these concepts of high-performance simulations, visualizations, and interfaces have been around for more than a decade. What’s really new here?
CM: The individual components have been explored or used one at a time. I don’t think anyone has really stepped back and [combined them] in the client devices and had enough computing power to do N at a time. We’ve taken handwriting or speech, or vision, and done a modest simulation of one of them, and figured out how to use it to navigate in an interface [people] already knew. A qualitative change will come in the next few years. When machines go up [in power] by a factor of 10 to 100, we’ll do several [types of natural interfaces] at a time. It’s a more transformational way of changing the man-machine interface. So these are areas I’m hopeful Microsoft, and its research activities as a company, will be able to bring forward.
X: How does all of this fit with Microsoft’s long-term global business strategy?
CM: Each generation of computing has been driven forward in long cycles by fundamental changes in hardware capability that leads to more powerful software capabilities, and then a more powerful way of using them. We tend to let a larger number of people work on a broader range of problems. Our fundamental business is software, and we are expanding the available market, in terms of more devices running more software.
[Developing nations] have to migrate their economy beyond a manufacturing base and into a knowledge economy. Europe and the U.S., to some extent, are both societies and economies that have benefited a lot from advances in technology, and find themselves in a much more competitive world.
X: Has anything surprising or different come up in your current discussions with students and faculty?
CM: Not really. Some things have been reinforced for me. Students have a fairly simplified view of information technology. They experience it on their laptops and cell phones. Their applications are dominated by social networking. It takes a real diligent effort by companies like Microsoft, and faculty, that students grow up with a broader sense of computing. In interactions with faculty in computer science departments, they exhibit a strong interest in these trends, but the curriculum is not evolving as quickly as the general field of computing is evolving. We tend to see the applications of computing as we’ve known it, instead of the future of computing.
X: How is the culture of Microsoft changing post-Bill Gates—if it is—in arguably the most challenging period in the company’s history?
CM: Cultures in big organizations don’t change with the wind. They evolve over a long period of time. They’re not static either. We’ve managed the transition to Ray [Ozzie] and me. I think that was done very gracefully, with no real disruption. It was not a material event in some cultural evolution. We are adapting to the environment, and adapting the business to the markets today. Those things are more of a driver to how the company adapts [than Gates’s departure].
Clearly the “cloud,” as everybody calls it—a very large-scale computing ability—is an important addition. We’ve been adding a service component to all of our products. I like to think we’re tracking these things, and we are diversifying. We’re way more diverse now, whether it’s in servers, entertainment, devices, software for small and medium businesses, energy, or health.
X: Talk about the significance of research to Microsoft’s future, as you see it now.
CM: The willingness to devote a percentage of our [money to] research is ultimately the key to the company’s health and survival. When we want to enter a new business like energy or health, we look to the research aspect as a component to getting those things started. Research is like a shock absorber that lets the company respond when the markets or [actions of our] competitors were not anticipated. It will continue to be important.
X: Search and online ads would be an example of this, right?
CM: The company waited too long to directly enter the search and ad business. Even Google, the key competitor there, when they started, they were just doing search. When it became clear that ads were an important coupling, Microsoft decided that was an important area. Going toe-to-toe in search has been aided by research. But in Bing, we needed to advance beyond the concept of search. That’s a good example. It’s an important part of how we’ve been able to compete in a core technology area, and add some novel capability.
X: In your travels, you’ve built strong relationships with foreign governments—which is crucial to Microsoft’s future as a global company. And I’m told you know Henry Kissinger, the former U.S. Secretary of State and a noted political advisor.
CM: Henry has been a personal friend for the better part of a decade. He’s a guy I learned a lot from. As I was trying to understand foreign policy, he was trying to understand technology. So he was my foreign policy coach, and I was his computer coach. We’re half a generation apart. When I traveled in China with Henry, he helped me understand a lot about the Chinese people and government. I’ve benefited from traveling around with him, as a trusted intermediary with many of the world’s governments. I look to him as a role model. A lot of my business is over the long term, and my investments on the relationship side transcend technology.
X: So has Dr. Kissinger tried Windows 7 yet?
CM: I don’t actually know. I doubt that he ran out to the store on October 22.