This morning's keynote speech started with one of those tingly-feel-good Microsoft Research commercials.  You know, the type of advertisement that really makes you believe that Microsoft can change the world.  This seemed very suitable, as the first speaker of the day was Rick Rashid, Sr. Vice President for Microsoft Research.

If you need to imagine the scope of Microsoft's Research arm, it houses over 850 PhD researchers in 6 different locations -- a larger faculty than the entire CMU, or Brown University.  This is the equivalent of making a Berkley Computer Science factory every year for 17 years without a single employee leaving or retiring.  Microsoft Research has over 4,000 peer reviewed publications over the past 17 years.  Mr. Rashid says "If you go to any major conference in the field of Computer Science, Microsoft will have between 10-30% of all papers at the conference."

It's not all just fun academia stuff -- much of the research from this group become products in Microsoft.  Need some example?  Many parts of Vista, ClearType in Windows XP, Natural Language and Speech in Microsoft Office, Windows Media audio/video codecs, the Tablet PC, Datamining in SQL Server 2000+.  There's also some new products -- take Microsoft Robotics Studio -- that are really, really cool and come out of MSFT Research.

So, what's next?  There's a continued push on Voice, Video (gesture), Ink and Tablet inputs.  As you dive deeper in to application layers, you'll quickly see that Microsoft's Research arm is trying to solve problems from data storage and optimization to new challenges posed by Cloud Computing.  But it's not just all technical challenges, some of the challenges Microsoft aims to solve with this research arm are related to compliance, governance and financial.  The role of Darpa has changed over the past 20 years, government grants aren't as available as they used to be and it's critical to maintain the drive for innovation.  Microsoft is committed to working on this problem as well, not just the technical challenges that can be solved because of this work.

We also saw some concrete examples of what Microsoft Research is doing now.  How about applying Bayesian analysis while analyzing the mutation rates of HIV strands -- you know, that same logic that's used in many popular span filters including the one in Microsoft Office Outlook? 

 
Perhaps the coolest thing I saw from Microsoft Research this morning was the Worldwide Telescope project.  This completely awesome 3D model analyzes thousands of gigabytes of data from telescopes and satellites all over to make comprehensive and compelling 3D models of the sky.  Take a tour of galaxies far away or our own solar system.

How about a sophisticated, interactive virtual landscape that teaches children the basic concepts behind software development?  Boku is just that -- and it runs on your existing Xbox 360.  This thing was awesome, and reminded me of a much cooler version of Logo, a simple programming language I learned as a young child.  I'll be putting my 7-year old nephew on this next year when it comes out.

And there was more.  The next version of Microsoft Surface, called SecondLight, contains 2 layers of interactive screens, one on the table surface, and one projected onto any porous element you put on top of it (such as a piece of paper or a plastic disc).  And then there's all that fancy stuff being done in robotics by this group.  Yes, 850 PhD researchers really can put together some pretty cool tech -- the type of tech that will likely change the world.