Microsoft does something not terribly evil

June 9, 2004 4:02 PM

Scientific American on Microsoft's new, gigantic attempt at a Palo Alto.
posted by Tlogmer (7 comments total)

Interesting, but a pretty lightweight article.
posted by Ethereal Bligh at 4:26 PM on June 9, 2004

It would be nice to see Bell Labs revisited. AT&T had to give it up after they lost their monopoly, but Microsoft still has its monopoly. Bill Gates has shown incredible generosity in charitable contributions, and in some respects a Bell Labs environment is such an endeavor. However, it also spins off lots of technology useful to its founders. In today's fast paced environment it may be even more valuable. Part of the loss for AT&T was the incredibly long lag between innovation and commercialization at the time. Perhaps today stellar advances could hit the market before the patent actually expires.
posted by caddis at 5:16 PM on June 9, 2004

neustile -- it may have started then, but it's only recently become a big enough effort to be important; MS has been luring in top comp sci people gradually (that was the impression I got from the article, at least).
posted by Tlogmer at 6:26 PM on June 9, 2004

Microsoft's problem isn't that they need more smart people working for them, it's that the universe of ideas they can exploit commercially is tiny. Their commercialization of research must by necessity be constricted to those areas that have absolutely no real impact on either operating systems or applications software. Microsoft knows how to sell Windows, and they know how to sell Office, and they are very, very good at it. If the world were to change appreciably in any way, they would risk disrupting their money-printing machine.
posted by Vetinari at 9:45 PM on June 9, 2004

The amusing thing about that article is that the exact same "research" ideas are shown every year at TechFest. Google around, and you will find the same "innovations" being shown in 2000, 2001, 2002. On the plus side, they were giving out 256 MB USB memory drives last year. If any of your friends work for Microsoft at the Redmond/Bellevue location, make sure they bring you one.

There are also some interesting technologies and tools that are developed in parallel in the research org and the product orgs. Microsoft doesn't have the software development discipline that Open Source mandates, so it is often faster just to write new code from scratch than to try to replicate the custom build environment that some random researcher is using.

It is cool that Microsoft hires a bunch of PhDs, gives them decent salaries, and all the Microsoft software they can handle. But don't expect either (1) MSR to have any positive impact on revenue or the stock price or (2) MSR to cure cancer* or otherwise contribute generally to humanity.

*cancer: Bill contributes to outside organizations that are trying to cure cancer. If MSR was so great, you would think that Bill would either ask THEM to cure cancer, or expand their ranks to encompase cancer researchers.
posted by Kwantsar at 10:59 PM on June 9, 2004

I wonder how many of these researchers are dedicated to security related topics?
posted by PenDevil at 12:16 AM on June 10, 2004

"They are a bigger and of course better-funded version of many other somewhat successful labs around the US; ibm, nec...": in fact I don't think Microsoft Research is bigger than IBM Research (the numbers that I find are 700 and 2,000+ people respectively, although I don't know if they're counting exactly the same things). I believe IBM also spends more on Research than Microsoft does (at least that's what I find on the Web).

[Note that I work for IBM Research; but I've never actually counted the people or the dollars myself... *8)]
posted by davidchess at 7:38 AM on June 10, 2004