Having failed to acquire Yahoo, Microsoft has been busy spending money, with news that they've acquired natural language search company Powerset for $100 million.
Powerset sort-of launched in May, with a smart search service that only indexed Wikipedia and Freebase entries. In use, Powerset is like the original Ask Jeeves on speed, focusing on questions as opposed to keywords. For example, a question might be "who invented RSS." Google returns various articles debating the subject, Powerset returns a list of actual results focused on the question. (I always thought it was Dave Winer, but apparently Netscape was involved, so I learned something).
The big question: can Powerset save Microsoft's Live Search?
Microsoft's online strategy has failed to capture anyone's imagination except for those people so computer illiterate they don't know how to change the default search engine in Internet Explorer. This is not to say that parts of Microsoft's online effort aren't really cool, Popfly, and Windows Live Maps are leaders in their relative spaces, but search has been a failure for Microsoft, irrespective of how much money they throw at it. And it's not that Live Search is really bad, it isn't (it's significantly better than Yahoo in my opinion), but it's not Google either.
The important thing to consider with Powerset is that it delivers a distinct point of difference to Google. Live Search powered by Powerset becomes Ask Jeeves 2.0, and although Ask Jeeves itself may never have been wildly successful, it was due more for a failing of its algorithm at the time then the concept of natural language search not having a place. The flip side: Powerset is not proven on indexing anything more than Wikipedia, so we don't know really how good the algorithm is. We can take that it does work to some extent by Microsoft acquiring it, but until we see it in the wild it's hard to make the call.