Why Microsoft will fail to do anything significant with Powerset

Why Microsoft will fail to do anything significant with Powerset(techcrunchit.com)

3 points by Aleran 17 years ago | 3 comments

jm4 17 years ago |

I am thoroughly convinced Microsoft will botch up this acquisition, but that's only because it's usually a safe bet that Microsoft will botch up anything internet related. They just don't get it.

This article is based largely on the assumption that priority number one will be to port everything to Windows and I think this is an absurd assumption to make.

Are we to believe that Microsoft is willing to risk throwing away $100M+ in order to migrate this thing to Windows?

Don't forget that Microsoft ran Hotmail on Unix for years. I imagine by the time they got around to porting it to Windows they were about ready for an upgrade anyway. (Plus, Windows was finally getting to the point where a project like that could even be considered.)

My guess is the meat of this deal was about bringing in some new IP and some bright engineers- not about running some search engine that almost no one has heard of. In that case, the platform it runs on is largely irrelevant.

Aleran 17 years ago | |

The difference between Hotmail and Powerset is Microsoft has said that they will have some sort of integration with Live search by the end of the year.

http://www.techcrunch.com/2008/07/02/interview-with-barney-p...

Maybe I don't know what I'm talking about but having half of Live search running off Windows search and the remaining part being handled by Powerset's unix boxes sounds silly. Or at least I will think I am seeing pigs fly if it every happens.

jm4 17 years ago | | |

I did see that Microsoft had said they will have some sort of integration by the end of the year, but that doesn't necessarily imply that they will need to port the Unix stuff to Windows. Maybe a mix of Live Search on Windows and Powerset on Unix sounds silly to some, but you could also make the argument that dumping a bunch of time and money into porting a working application after a $100M acquisition is silly.

The article you linked to mentions that only 5% of searches contain natural language elements. This integration they talk about could be as simple as identifying these searches and passing them off to the Powerset stuff in the background.

I don't doubt that Microsoft will eventually want to port everything to Windows, but I think it's a false assumption that this is going to be a top priority. They've made repeated statements about their desire to see this put to use quickly and speculation about a lengthy and complicated port seems to directly contradict that.