(In case you missed it, here’s a profile of Craig Mundie, Microsoft’s chief research and strategy officer, written by my predecessor, Sharon Pian Chan. The story ran in the print edition of The Seattle Times April 1. – Janet I. Tu)
As Microsoft’s strategic thinker, Craig Mundie prefers to live on the bleeding edge. So when he and his wife moved to a house in Washington Park, he wanted to trick it out with servers to control the lights, the computers and entertainment systems.
His wife just wanted a television she could turn on with a single button. “She said, ‘Well if it’s all so highly automated, I want it to be like the good old days. I pick up the remote control, I hit one button, I get to watch television,’ ” Mundie said. He gave in. She got what she wanted.
Still, Mundie is pushing Microsoft toward a future without buttons — no keyboard buttons, mouse buttons or phone buttons.
Touch-screens are just the first step, Mundie said. In his version of the future, people will talk to computers without having to type on a keyboard. The computer will understand hand gestures and body movements instead of mouse swipes. He calls this evolution “natural user interface,” or NUI for short.
Computers will become more like humans, he said, able to understand such language as, “Where should I eat tonight?” instead of forcing people to type the disjointed words “best,” “restaurants,” “Seattle” into a search engine.
“I’ve been the biggest, steadiest champion inside the company for that transition,” Mundie said.
As chief research and strategy officer at Microsoft, Mundie oversees the Microsoft Research group. He also incubates technology startups within the company and works with government leaders around the world on technology policy.
Mundie has already helped nudge video games toward natural interaction with Microsoft’s Kinect, the Xbox motion sensor that replaces the hand controller, which has been around since Pong. With Kinect, a player throws an imaginary ball at the television screen, and a character in the game hurls a virtual fastball.
The Kinect was the fruit of years the company invested in Microsoft Research, a group of 850 scientists who conduct academic research, a blue-sky pursuit in a company otherwise focused on profits. Microsoft spends $9 billion a year on research and development. The number includes product development, but it is larger than the National Science Foundation’s $6.9 billion budget last year.
The work of Microsoft’s researchers has found its way into many products at Microsoft, but Kinect was the first consumer hit. “This is the first mass-market example,” Mundie said. “We threw everything we had at that problem.”
To develop the technology, Microsoft researchers worked directly with the Xbox business team in a project code-named “Natal.” The first prototype took four months and $30,000 to build. It’s now priced in stores at $149, selling 18 million units last year.
Mundie had been thinking about combining the television with the computer for 15 years. In 1997, he helped lead a Microsoft investment of $1 billion in Comcast. The same year, he worked on Microsoft’s $425 million acquisition of interactive TV company WebTV.
He first joined Microsoft in 1992 to create and run Microsoft’s Consumer Platforms division, which built a version of Windows called Windows CE to run devices other than computers.
In his two decades at Microsoft, Mundie has served, among other things, as chief technical officer, working with Chairman Bill Gates to develop Microsoft’s strategy around global policy issues.
He continues to jet around the globe in his current role, but Mundie remains more geek than international man of mystery. Born in Cleveland to an accountant father, he grew up in Detroit and New Orleans. He taught himself to use the slide rule in the 6th grade, only to find it broken one day by a class bully. Even now it’s hard for him to believe that technology is considered cool and people like Mark Zuckerberg and Steve Jobs have hipster fan clubs.
Mundie acts as a Henry Kissinger for Microsoft, advising different countries about fostering tech innovation. He serves on President Obama’s Council of Advisors on Science and Technology. It’s his diplomacy skills that helped bring the business and the research sides of Microsoft together.
Don Mattrick, president of interactive entertainment at Microsoft, said he was struck by Mundie’s willingness to blend the two in the development of Kinect.
It was “an example of where one plus one equals three,” Mattrick said. “The business group … would have had a hard time executing on the research for Kinect. The research team didn’t have the manufacturing muscle of the business group.”
The research and development has continued since. Microsoft has launched a holographic virtual-meeting software that uses a player’s Kinect avatar. Engineers programmed the Kinect to recognize and replicate facial movements, such as smiles, scowls and raised eyebrows.
In February, Microsoft began selling Kinect for Windows so developers can add motion sensing and voice recognition to computers.
PC as assistant
Making a computer understand how humans interact is just the start, Mundie believes. The next step, he said, is getting the computer to behave like a human.
“The computer will become less of a tool, more of an assistant,” he says. “You know — book your flights, make your hotel room [reservations], arrange your meetings, manage your calendar.” The computer can do that now, but a person still has to steer it by going to a travel website, adding appointments in Outlook and sending emails.
“You want to be able to go to the computer and say, ‘Schedule a meeting with Sharon,’ ” and the computer automatically knows to search for a meeting time and schedule a conference room, Mundie said.
Mundie’s message of the computing future echoes across Microsoft. Chief Executive Steve Ballmer espoused the same vision in multiple speeches last year.
Mundie draws an analogy to the auto industry.
“We have built some fantastic cars. They have become more and more powerful. But you still have to get in the car and drive it to get to where you want to go,” he said. “The dream is to be able to get in your car and tell your car where you want it to take you. And there’s a huge leap between those two things.”
The question is whether Microsoft can make products beyond the Kinect that will dominate this future. The company was an early investor in many technologies that other companies now lead. For instance, it added touch-screen technology to Windows with the Tablet PC long before Apple came up with the iPad.
“Touch has been a part of Windows for a very, very long time, but for whatever reason they couldn’t manage to get a whole bunch of us using it,” said Michael Cherry, an analyst at independent research firm Directions on Microsoft in Kirkland. “When people think of touch these days, I don’t think they think of Windows 7. They think of the iPad and iPhone.”
Cherry sees a similar failure with voice technology.
“Microsoft has been doing voice recognition for a long time,” he said. “But if people were to say to me, computers using speech, [Apple’s] Siri would be the example today.”
Even Mundie’s example of a self-driving car has already been accomplished by a Microsoft competitor. Google built a prototype in 2010 that drove from Mountain View, Calif., to Hollywood.
Decades of research
Mundie says it takes more than two decades for research institutions to hit their stride, and he wants to find more ways research can spark blockbuster products like Kinect.
“When you look at many of the historically great research institutions, it sort of takes them a decade to get going, get warmed up, a decade to start to produce fundamental new research,” he said. “I’d say Microsoft, through the years, has been at least as good as anybody else ever was, and right now I think we’re on a path to try to be better than any one of the major labs has ever been in terms of steadily affecting products.”