It sounds like the next step in search might be audio and video.
Oh, sure, everybody is working on those things, but just take the idea of finding your local pizza place and doing that right; search doesn't do that well today. Search is really crummy today -- it's just that it used to be really really crummy, and now it's better, and there never was anything like this before. So most of the results people get back today are irrelevant results. Deep analysis can take us much further, and that's why we're investing a lot, and you'll see us more very rapidly.
 
This year, there is a big push to make cheap computers for emerging markets. How is that going to have to evolve? A $300 computer is still going to be too expensive for many, probably, in Russia, India and other places.
Well, that's not really true. The expensive thing is the connectivity. Getting Internet connectivity is expensive. If all they had to do was pay for the computer -- $300 -- and the communications were free, then we'd see that PC usage would be very, very big. Ironically, communications costs tend to be highest in developing countries.
 
So you need to do shared PCs through cafes or community centres. We're doing a lot of so-called mesh networking research that could use wireless spectrum to try to bring those connectivity costs way, way down, and that's where you really see widespread computing breakthroughs.
 
The hardest thing, by far, is communications, then the hardware costs. We can make sure the software cost is never really holding things back -- that it's a small enough percentage. In educational things, we do a lot of software giveaways. We've been very generous to make sure that when people first come into computing, software doesn't hold them back.

That's interesting - the mesh - because that way, people can actually share a connection.
Yeah, the hard thing is the back haul of the Internet out of the village. Within the village, sure, you can mesh that up, but if people are going to be streaming video, you need quite a bit of capacity there. So it's not a simple problem to solve; we've got actually multiple research locations at Microsoft working together on this mesh thing, and we've had a lot of conferences working with third parties, so we're optimistic that that's the thing that can solve the thing that holds back developing-world computing.