Posted by: jonkatz | December 16, 2009

Hiring at MSR

In recent news, Microsoft New England has apparently hired Boaz Barak. (Though I could not find any announcement by Microsoft, or any indication on Boaz’s webpage.) This comes on top of Microsoft’s recent hires of Madhu Sudan and Omer Reingold (in Silicon Valley). One could argue that were MSR New England an academic department, it would rank among the top 3 in theoretical computer science. (Of course, one could also argue that were MSR New England an academic department many of the people would leave and the productivity of those who remain would be cut by half…)

I have to admit to feeling a bit jealous, as MSR New England is one place I would love to go… (and yes, I’ve “applied”).

I don’t want to re-hash the debates about academia vs. industry. (Anyway, it is not a fair comparison in any of these cases since Boaz, Madhu, and Omer are getting the benefits of industry without having to give up the benefits of tenure.) I’m more interested in the question of what Microsoft expects to get out of these hires? Or, more bluntly, how does Microsoft expect any of these hires to impact its bottom line? I am not asking about the value of TCS research in general, which I think is clear. I am asking, in particular, about hiring people who are about as far from practice as possible. I don’t think comparisons to “the old Bell labs” work here: the fundamental physics work done at Bell labs seems, to me, much closer to yielding practical benefits than fundamental work on complexity theory. I also don’t buy the arguments about these hires being public relations coups: how many people, outside of the attendees of STOC/FOCS, would recognize these people’s names? And for how many people will this affect their decision of whether to apply to Microsoft for a job, or buy Microsoft products?

But maybe I’m missing something. Any thoughts?

Advertisements

Responses

  1. This probably sounds very biased, but I think STOC/FOCS over represent complexity theory, thus making complexity people appear to be the most brilliant and thus the most sought after.

    An algorithmic idea that could be useful in practice is unlikely to receive such attention, thus making people of this genre not as sought after.

    There are exceptions to what I am saying, but this is a general trend.

  2. I am extremely impressed at their ability to hire senior people. MSR Silicon Valley seemed for awhile to be hiring lots of junior people. But getting a job at New England just seems completely impossible, when the recent hires are Madhu Sudan and Boaz Barak — and even the more “junior” people like Adam and Yael Kalai were hired away from faculty positions at top-10 theory schools.

  3. “I’m more interested in the question of what Microsoft expects to get out of these hires? Or, more bluntly, how does Microsoft expect any of these hires to impact its bottom line?”

    I think, this question is similar to ask a question, what benefit the IBM is expecting by pursuing cryptography research in Watson research center?

  4. With all due respect to Madhu, Boaz, Adam and Yael, do you really think 4 people can make a top 3 theory department (even restricting oneself to the US)? Am I missing someone? I

  5. Also, MSR probably wants to “become part of MIT” which will help their reputation and possible their bottom line. They sponsor many well-established MIT seminars and put their name on them.

  6. In the mythical good-old-days companies would do things for the good of society even if they did not help the bottom line. I wonder if some of that mentality is here, like a law firm doing pro-bono work.

  7. I think, this question is similar to ask a question, what benefit the IBM is expecting by pursuing cryptography research in Watson research center?

    Not it’s not. Cryptography has direct applications to practice while complexity, for the most part, does not. I am not suggesting MSR should not hire complexity theorists (on the contrary I think its great) but just pointing out that Jon’s question is not similar to “why IBM pursues crypto research”.

  8. I think, this question is similar to ask a question, what benefit the IBM is expecting by pursuing cryptography research in Watson research center?

    I’ll just add that not only does the work done by the crypto group at IBM have the potential for applications, but in fact the crypto group at IBM actually spend some of their time working on real products and applications. Some of this work is publishable, but much of it is not. But it’s very valuable to IBM.

    I don’t get the sense that this is happening at MSR New England (though it does happen at MSR SV and MSR Redmond).

  9. how does Microsoft expect any of these hires to impact its bottom line?

    It isn’t just that this does not impact MS bottomline, it has an incredible opportunity cost for MS. Given the same resources, one can only imagine the impact that could have been if MSR hired the “right people”.

  10. actually, what i was told (during an MSR redmond orientation session last summer) is that MSR is the only industry lab that does not have company-directed research, i.e. they support fundamental research without regard for the company’s bottom line.

  11. Re the last commenter’s quote about MSR propaganda. That seems like a complete bogus statement given the experience of many researchers at MSR who are not working in theory.

  12. What I don’t get is that if Bing is going to overtake Yahoo’s search business and compete with google, why are they not concentrating on hiring algorithms people to improve their search engine?

  13. Strong and long term connections with academic departments can be one motivation behind hiring strong and influential people. While these hires might not get noticed by people on the street, they will get noticed by bright undergraduate and graduate students who will be on the job market soon!

    For example, I know of several former interns at MSR that went on to accept non-research positions at Microsoft.

  14. Amen.

  15. After reflecting a bit more, I agree that there is both company-directed and non company-directed research going on there. So I guess my point is that MSR does (but not only) support fundamental research without regard for the bottom line. To what extent someone does one or the other is probably position-dependent. This seems to explain what is going on in New England. (And, speaking from my own experience, I was given complete freedom during my internship to work on whatever I wanted.)

  16. The last comment was from me.

  17. I work next door to MSR Cambridge (UK). There was lots of suspicion when MSR was founded, because they were originally to have the top floor of our lab and there would be all sorts of conflict of interest issues (what you could say in the cafe, because people from MSR might overhear it). And there was a feeling that MSR was trying to buy its way into Cambridge University.

    I think having a separate building is a good idea. Largely they keep themselves to themselves, but come over to us for seminars etc (and we go over occasionally). The contributions from MSR people are appreciated. Indeed some people in MSR still have major teaching commitments in the University and elsewhere.

    I think an appreciation has grown that the MSR people are doing interesting stuff and are generally left to get on with it without too much interference from management. ‘M$’ have a terrible image, and MSR is going some way to rebuilding that in the academic community.

  18. Jon, would you move to MSR New England at your current_salary+adjustment_for_the_place?

    (Just in case, its a joke)

  19. 1. MS has a surplus. each new $ coming in will be divided among MS and the US government. Paying people money is a good way to “lose” money (in other words, the IRS pays some of the salary of these people).

    2. Why AT&T had their labs? ‘coz when you put smart people in the same place, and allow them to do whatever they want, once in a while they’ll do something amazing. And it will be so amazing, that it will be worth zillion of $.

    3. ‘Coz they can. I mean, would you rather them spend the money on a new campaign to use WinXXX?

  20. Boaz Barak now has a notice about this on his page, so it is confirmed

  21. Hi it’s me, I am also visiting this web page
    on a regular basis, this web page is really good and the
    users are in fact sharing nice thoughts.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Categories

%d bloggers like this: