Wednesday, October 15, 2014

You'll be fine in Silicon Valley

Yesterday, a group of well-known professors from a bunch of universities sent an open letter to Microsoft Research's leadership, condemning them for closing the Microsoft Research Silicon Valley lab and leaving the researchers there "jobless". Although I know and respect many of the authors of this letter, I think they get it wrong when it comes to the career opportunities for the Microsoft researchers impacted by the lab's closing, and more broadly for the impact this closing will have on the research community.

The letter's tone is incredibly haughty -- a mixture of "how dare you" and "think of the children!!" What worries me the most is the message the letter seems to be sending to junior researchers who may feel that industry jobs are not a viable career path, because, hey, you might get fired if your lab closes. I think we should correct some of these misconceptions.

(Standard disclaimer: This is my personal blog and the opinions expressed here are mine alone, and most certainly not that of my employer.)

Let's take the letter point by point.

By now, you are no doubt aware of the research community’s shock and disappointment at the sudden and harsh way in which the members of the Microsoft Research Silicon Valley lab were dismissed a few weeks ago.
While the news was sudden, I don't have any evidence that it was "harsh". That's how layoffs sometimes work. Yes, it sucks, but don't imply malice where none is intended.

We are writing to share our perspective on the negative impacts of the shutdown and to open a dialogue about the ways in which Microsoft can try to restore the environment that enabled MSR to produce such great research in the past, benefiting both the company and the world.
The implication here is that by closing the Silicon Valley lab, Microsoft has damaged "the environment" that enabled them to "produce such great research". Let's see here. By my count, about 75 researchers out of a worldwide organization of "more than 1,000 scientists and engineers" were affected. While this is unfortunate, I don't think it's going to seriously hinder Microsoft's ability to do great work in the future. It's still a great place and I expect it will continue to grow.

While layoffs are always unpleasant, the impact of this one has been exacerbated by the fact that many researchers at the Silicon Valley lab worked on long-term, fundamental research of the kind that can be done at very few places outside of academia.
The fact that "very few places outside of academia" allow one to do "long-term, fundamental research" should be teaching us something. The authors of this letter seem to believe that doing academic-style research is some kind of special, protected occupation that people with PhDs are entitled to. While there may have been plenty of industry labs doing pure, academic-style research 20 or 30 years ago, the world has changed. What worries me is that professors still cling to an industry research model that has largely been supplanted by other, better models. I think it's time to evolve our thinking about this in the community and start giving appropriate career advice to students so they are prepared for the reality of what's out there, rather than for an outmoded academic ideal.

It is also a common misconception that one cannot do research in a more product-oriented industry setting. For example, Google does a tremendous amount of research, albeit using a fairly different model than places like MSR. The same is true for other large and small companies as well. We may not call all of these places "research labs" and bestow fancy titles on people like "research scientist", but the research is still happening, and having tremendous impact.

As you know, the academic calendar is such that many of these researchers, including very junior ones who recently completed their PhDs, could be jobless for nearly an entire year. We feel that there should have been a better way to close down this lab, one that would have allowed them to have continuous employment until academic jobs are available again in September 2015. Given that this lab was continuing to produce exceptional — indeed revolutionary — research, we fail to understand why closing it had to be done so suddenly.
This is completely disconnected from reality. First of all, where did anyone get the idea that the only viable job opportunities for the researchers being let go from MSR SV are in academia? From what I can tell, all of these people are being heavily recruited by a number of Bay Area companies. Hell, the MSR SV office is literally off of the same highway exit as the Google headquarters. About the only thing that might change for some of these people is which parking lot they drive to in the morning.

I very seriously doubt that any of these researchers will be "jobless" for any length of time. That's delusional.

Second, as a friend of mine pointed out, it's not as though the "academic calendar" is some kind of law of physics. If universities are serious about recruiting some of these people, I am sure they can find a way to make it happen, even if they have to resort to such lengths as hiring them on short-term contracts as visiting scientists or whatever. The academic calendar is your problem, not Microsoft's.

Over the past two decades, MSR, and indeed all of Microsoft, earned an excellent reputation in academia as an organization that not only valued basic research but also supported the career development of the many researchers that worked in or visited the labs.  That reputation has been significantly damaged, threatening Microsoft’s ability to recruit and retain world-class researchers. 
It's true that shuttering MSR SV will have some impact on how people view Microsoft Research as a career choice. On the other hand, it's not as though they hired a huge number of people every year -- the total number of jobs impacted is relatively small. In comparison, Google hires multiple MSR SVs worth of people every week, and more PhDs per year than all of MSR employs, worldwide. (For that matter, Microsoft, as a whole, no doubt beats Google's numbers.) It's not like the loss of a small fraction of MSR's total headcount has anything but a minor impact on the overall pipeline for computer scientists.

What's really happening here is a bunch of finger-wagging in an effort to publicly shame Microsoft for what was no doubt a very difficult and complex business decision.

As faculty members, we can no longer recommend it as highly to our students as a place to start their careers.  In the long term, this move seems likely to adversely affect Microsoft Research (and the positive contributions it makes to Microsoft as a whole) in more ways than any benefit it may have had.
I think Microsoft should appoint a panel of academics to run the company, since clearly they know more about running a business than MSR leadership does.

Nevertheless, we believe that Microsoft can reduce the damage that has been caused by the shutdown of the Silicon Valley lab.  We understand that Microsoft is considering ways to help care for the researchers who were dismissed, such as defraying the additional costs of the academic organizations who are trying to provide these researchers with temporary homes. This would be an excellent, and highly appreciated, first step.
Of course universities must bear the cost of supporting these poor refugees while they find new jobs! I mean, it's not like the Microsoft stock and severance package will pay the bills for very long until the affected people get jobs at Yahoo or Twitter. Hopefully the universities can pull through with basic rations, perhaps a tattered wool blanket and a cot to sleep on, an office with an aging workstation running XENIX, just until these people can find gainful employment. It would be a real humanitarian disaster otherwise.

Looking forward, we hope that you will open a discussion with us and the community about Microsoft’s vision for industrial research (which has become less clear after the closing of what appeared to be an extremely valuable and successful lab) and concrete commitments MSR can make regarding the career development of its remaining and future researchers. 
Frankly, I don't think that the MSR leadership owes the academic community any kind of explanation or commitments whatsoever. If anything, this was a great wake-up call to anyone who was living under the false impression that being a "researcher" at a company gave you job security. This is how companies sometimes work.

Fortunately, especially in Silicon Valley, there are literally hundreds of great companies that the MSR SV folks can work for, doing all kinds of awesome and groundbreaking work. The tech industry as a whole is going strong, and I have no doubt that there will be all kinds of career opportunities for these folks.

Steps like these are essential to rebuilding the relationship between Microsoft and the academic community, along with all the mutual benefits that it brings.
On the whole, my guess is that the relationship between Microsoft and the academic community will continue to thrive. There are a lot of great people still at MSR and the ties there are incredibly strong. It's too bad that they had to shut down the Silicon Valley lab in the way that they did, but I'm not sure it's productive to cast a dark cloud over everything MSR does and stands for as a result.

Mostly, I'd like the authors of this letter to think about the message they are sending to students and junior researchers, which seems off base to me. We're all sad about the loss of MSR SV, but I think this letter really goes too far when it claims that anybody there will be "jobless" or that academic positions must open up to absorb the affected researchers.

70 comments:

  1. Out of the many, many, many holes one can punch in your manifesto here, let me just say the following two.

    (1) "Professors still cling to an industry research model that has largely been supplanted by other, better models." Better how? Better why? By what standard? Matt's scale of ranking research? One potential standard could be of sharing ideas and publishing novel results in top-notch conferences. According to that standard, MSR has a significant lead over other companies with so-called-better research models.

    (2) "Where did anyone get the idea that the only viable job opportunities for the researchers being let go from MSR SV are in academia?" Here is where you completely fail to understand the nature of research in MSR. It *is* the last industry workplace that conducts true academic research. Should the fired MSR SV researchers want to keep on pursuing their own research agenda (which they have worked on for years), keep their passions and interests and what they have centered their academic (yes, academic!) lives around -- then yes, they must leave for the academia. Moving to the Google parking lot means more than the few extra miles. It is shifting to a completely different mindset.

    Now, you may want to argue whether or not any company should allow for a bunch of (incredibly brilliant) researchers work on their own agenda. That is indeed debatable. But regardless of your own opinion, Microsoft, in the past, has answered that question on the affirmative and established such a lab. All that the letter asks for is that it continues showing these researchers the same respect it showed them when recruiting them --- and give them sufficient time to find an alternative place for themselves and their ongoing research.

    ReplyDelete
    Replies
    1. (1) Well, it's my blog, so for the purposes of this post, yes, by Matt's scale of ranking research :-) If the MSR model were so successful, why is it the last place that does it?

      (2) I agree that if they want to do academic-style research then they will likely have to go to academia. The point is that the letter says that these people will be **jobless** otherwise. Unless this is some new definition of "jobless" that I am not familiar with (and I'd love for some of the professors who wrote this letter to go to a place like San Francisco with lots of actually jobless people for reference), it seems quite likely to me that these people will be able to find gainful employment in other career paths. I would argue many of those paths will turn out to be just as rewarding as the academic path.

      Delete
    2. This. Leslie Lamport would make a hell of a SRE.

      Delete
    3. I think the way the letter is using the word jobless makes a lot of sense. The people in question are academic researchers, and unless they want to change their line of work and start new careers at the ground floor they are very likely to be jobless for about a year.

      By your definition of jobless, no one is ever jobless since anyone can find an entry level job (in the service industry for instance). You don't actually believe that the word jobless was invented to describe an empty set of people?

      Delete
    4. I hope you are kidding, since otherwise you appear to be completely unaware of the reality that is unemployment in the real world -- even for highly skilled people.

      Delete
    5. "I hope you are kidding" or "you appear to be completely unaware" - these are generic comments that don't address the issue at hand.

      Most highly skilled jobless people could take an entry level job in the service industry, but they don't because they are trying to maintain / build their career. Or because they wouldn't be able to support their family on it, so it would be counterproductive at that point.

      Most low-skilled jobless people are there because they can't get the service industry jobs (they have criminal convictions / don't have a car / etc).

      There's a complex structure to unemployment, and you are missing it by simply drilling down on the word "jobless".

      Delete
    6. "Anyone can find an entry level job" - this is a hilariously inaccurate statement. In fact, it's just inaccurate, there is nothing hilarious about it.

      Delete
  2. Interesting and thought-provoking points. Judging by what I've read about the computer industry over the years, it seems to me that the reasons why companies have been moving away from the Bell Labs/Xerox PARC model of labs that focused on long-term research that may not lead to foreseeable products to either applied labs like the modern IBM and HP labs that focus on research that has a high chance of becoming part of a product within the next two to three years, or the Google model where research is integrated into the software design process for some of its products, is because the latter models have been achieving a higher "bang-per-buck" ratio for productizing the companies' research efforts. Some companies, like Apple, don't even have research labs and are mostly invisible from the research community. Companies don't start research labs necessarily out of charity (although some do); they start research labs to find out ways they could remain on the right side of the technology curve. Even with the famous Bell Labs, one important reason why AT&T invested so heavily in Bell Labs to begin with was because the federal government required them to in order for AT&T to legally maintain its telephone monopoly. I don't think AT&T would have invested so heavily into Bell Labs if it weren't for this agreement with the federal government.

    Even though there is exciting work going on in the more product-oriented labs and development teams in the Silicon Valley, I am still concerned about the long-term future of long-term research. Since industry for the past 15-20 years has increasingly been shifting to short-term, applied research, this is leaving academia and government to be the only places to do long-term research. But academia and government have their own pressures. For assistant professors, there is the pressure to publish prolifically in areas that are interesting to the research community; an assistant professor who publishes in his or her area of interest without regard to community interests and trends runs the risk of being unpublished and thus denied tenure, which results in needing to exit academia. Even when a professor earns tenure, he or she is still required to publish prolifically (which means keep in line with the interests of the community) to avoid departmental sanctions. These pressures may conflict with the professor's desire to do research in his area of interest. I believe you covered this in your blog in an article discussing academic freedom. As for government, while I hear the research environment is fine now there, there may be growing budgetary pressures in future years as the national debt continues to increase (my discussions about debt and the future of this nation are another topic).

    My concern is that if there are fewer institutions doing long-term research, there may be less of a chance of radical, revolutionary ideas being created that have no immediately obvious products in mind. The development of GUIs like the Mac and Windows, for example, is based on the research results from Xerox PARC. The Mach project at Carnegie Mellon and the Sprite project at Berkeley led to major influences in future OS and file system design.

    Maybe the solution isn't to return to the "good old" days of Bell Labs and Xerox PARC (and I wasn't around during those days; I'm just a grad student, so maybe my perception of those labs is rose-colored). Hmmm, maybe long-term research is a charity after all and should be funded as such (maybe we'll have the Mark Zuckerberg and Priscilla Chan Foundation for Computing Research one day). Maybe there'll be a computer science equivalent to the Institute for Advanced Study for mathematicians. But I agree that it is very important for young researchers like myself to understand the realities of the computer industry and how research is funded today, and to learn how to have meaningful research careers in this environment.



    ReplyDelete
    Replies
    1. You make excellent points and I can't disagree with anything here. I've worried a lot about the need for long-range research as well, which is one reason I think MSR is a very important place. Still, the world is pretty different than it was during the heyday of Bell Labs and Xerox PARC, and I think our models for training students have to adapt to that.

      Delete
    2. Just saying "different" avoids any sort of qualitative judgement. Is the world a better place without PARC and Bell Labs in their respective heydays? How much of what we're doing now is feeding off their seed corn?

      That's not to say that their models are practical in today's environment (Xerox had revenue of over a *trillion* of today's dollars, they could afford massive speculative bets) but just because there is less money to go around today doesn't mean that investment in pure and highly speculative research isn't good for everyone.

      A more interesting question (to me at least) is how we rehabilitate and encourage pure research. If individual companies can't afford massive investments in pure research and funding cycles mean that Academia can't engage in as many speculative bets, how *can* we make it possible for people to explore these problems?

      Delete
    3. Well, start by voting people into Congress who will actually support the NSF, and find ways of getting Lamar Smith off of the House Committee on Science, Space, and Technology. If industry isn't going to do pure research, then you need the government to fund it.

      http://news.sciencemag.org/policy/2014/10/battle-between-nsf-and-house-science-committee-escalates-how-did-it-get-bad

      Delete
  3. Wait, you're telling me that some professors have a massively inflated sense of self-importance and are out of touch with reality? I have never heard such claims before!

    ReplyDelete
    Replies
    1. I think if you step outside the software engineering world, you will see that the reality is basically the opposite of this.

      Software engineering is captivating to many people as a teenager / young adult, but it makes for a terrible career.

      * The education/training needed to become a computer programmer is minimal. Basically, a couple of years of college courses is enough.

      * Computer programmers get worse as they age. By contrast, a doctor or professor gets better with age pretty much into their 60s. For instance, a 40-year old computer programmer is already considered "old", while a 40-year old doctor or professor is considered "young".

      * Computer programming is easily outsourced. Not so much for doctoring (or policing, etc).

      * Computer programmers mostly do not interact with the final customer and are interchangeable. Sure, when you learned to code as a teenager, you thought your understanding was unique. But when work on a large system at a large company, you are pretty much a standardized unit of labor.

      None of these are the fault of computer programmers - they are just structural realities.

      Because of these terrible realities, most people who end up in software pivot into something else. The smart people will pivot in their 20s (medical school, Wall St careers, etc). The less smart (or less lucky) will be left to pivot under pressure in their 30s (including trying to make it in management or product design).

      The decidedly less smart simply get angry whenever they are faced with the reality that people in other professions enjoy longer careers / better career ladders / more job security / etc.

      This is why you can always find a computer programmer (or service employee, etc) calling for doctors (or nurses, government employees, professors, schoolteachers, etc) to be subject to worse working conditions.

      Delete
    2. I forgot:

      * Computer programmers have little control over their work product. Major project can get years behind schedule because of poor upper-level management. Someone's pet project that is doomed to failure can be dragged along for years. Highly worthwhile projects can be easily canceled because of politics.

      Delete
    3. I disagree with almost everything in this comment, so I simply don't know where to begin. I see absolutely no data to back up any of your claims.

      Delete
    4. Again "I disagree with almost everything" is just a generic answer that doesn't address the issue at hand.

      Let's take a specific point - a 40 year old computer programmer is considered "old". If you are 40 and still working as a computer programmer (as opposed to a manager or designer), you will have trouble getting new jobs and will feel pressure at your current job.

      No-one will even think of a 40-year old doctor or professor as old.

      The exact age doesn't matter - it could be 40, 41, 45. Do you disagree with this?

      Delete
    5. It very much depends on where you're working - is it a startup ? what kind of startup ? is it a big company? etc. Additionally, let me assure you young people feel "pressure" at their jobs as well. Furthermore, becoming a manager may help sometimes but it has the effect of removing you from directly technical work and many, many engineers prefer to stay as technical as possible. Lastly, the idea that a designer has an easier time getting a job as he/she gets older than a software engineer is .. bizarre.

      Delete
    6. It's not really because of a specific company. It's human nature that a computer programmer starts getting less effective some time after 35 (maybe it's 38, maybe 42, the exact age doesn't matter).

      Because of this, many employers (Google, startups) will not hire older programmers and will pressure existing people to get promoted or get out as they age. Sure, out there somewhere (municipal governments? university IT departments?) there are employers that don't do this, but since many employers do, the overall industry effect is the same - the funnel narrows.

      For, say, doctors the nature of the profession is different - people get better at it as they age. A 25-year old doctor is still a medical student, and a 35-year old just finished residency. At 45 a doctor is better at it than at 35, and at 55 better than at 45. The same is true of many other professions.

      With designers, I don't mean just graphic designers, and I am sure they have their own challenges. But you will often see people transition from programming to design in their 30s. I've never seen a designer transition to programming in their 30s.

      Delete
    7. My experience and that of people I know is so different from what you're describing that I think we're just exchanging anecdotes at this point. Google will not hire older programmers ? That is not so. I think it is probably more difficult to be hired over 50 but a 40-something software engineer who is good will absolutely find a job. The notion that a "computer programmer starts getting less effective after 35" - not sure if you believe that or are stating a perceived bias. If you believe that, we're arguing from very different starting points. I do agree that the software engineering field is more enthralled with youth than say the medical field, but the notion that your life ends at 40 in software engineering is just not true. Unless you're using a very narrow definition of software engineer - absolutely, many engineers transition into 'architect"-type roles (if that's what you meant by "design"). It is true they will not be "entry-level" engineers forever, but not sure why that's abnormal. Also, it's simply not true that a couple of years of courses make one into a computer programmer. You can maybe get a job coding something (anything), but there is a huge difference between "any coding job" and "software engineer" with any future in industry.

      Delete
    8. To be clear, i am not arguing being a software engineer is "better" than being a doctor (though doctors have their own challenges). It is a different, newer industry with a completely different labor pool (e.g. including non-US workers, etc.) and career paths are not as linear as those of doctors or teachers. However, that is not a bad thing by itself - but if people want long-term security and the same job until they are 70, being a software engineer or an industry researcher is obviously not the right job for them.

      Delete
    9. Anon @ 2:55pm - actually my anecdotal experiences agree with what you are saying.

      It only by looking at statistics and connecting my experiences with statistics that you start to see the big picture. For instance, does Google have employees over 40, and over 50? Sure. Does Google hire people over 40 and over 50? I am sure it does.

      But the median age at Google in 2013 was 29. On the other hand, every 29 year old (Google employee or not) is going to be 44 in 15 years. What does this mean *statistically* for older people trying to get a job at Google or trying to stay at Google?

      You could say that Google is a young company, but Microsoft has a median age of 34, and it's been around for a lot longer than Google.

      Delete
  4. Dude, do you really think that Omer Reingold is going to publish his next breakthrough on logarithmic space while employed at Twitter?

    ReplyDelete
    Replies
    1. Precisely. Matt is missing the whole point. So I'm going to state the obvious. For both systems and theory people at MSR-SVC, this is devastating news for the following reasons.

      (1) Each employee at MSR-SVC had their own research agenda. The job was like that of a professor at a top university (except that they didn't have to write grant proposals, which even people like Matt hate, and they didn't have to teach classes).

      (2) If they join Google or Twitter, they can no longer work on a research agenda that they personally set (this is true for both systems and theory folks, and worse for theory folks since the only place where they can do work is in academia).

      (3) Within MSR, SVC lab was extremely unique (this is based on my understanding interning there several times in several of their labs): other labs are more like group-based work (i.e., a lot of people come together and work on some problem). In SVC, there was more autonomy and resembled the job of a professor (with differences noted above).

      Delete
    2. I'm sorry, but that's not what the letter says. The letter says that these people would be *** jobless ***. That is what I'm objecting to. I fully admit that the kinds of jobs that may be available to these people may not be pure academic style research, but that is vastly different than being out of a job entirely.

      My point is that the letter is very out of touch with the reality of the job market, especially in Silicon Valley.

      Delete
    3. Er -- yes, they could get jobs if they wanted to work in a different area (say software engineering rather than CS theory), but that is true of many "jobless" people: out of work software engineers in Nebraska might be able to get jobs at McDonalds.

      What is the point exactly? If they want to stay in the same field, they have to go on the academic job market now and be without a paycheck at least until August. This is not true of the software engineers who were laid off and want to stay in the same field (software engineering), and is precisely why the layoffs for MSR should have been done on a different schedule.

      Delete
    4. See, I'm just not buying the premise that any of the MSR SV researchers, even the most theoretical of them, are forced in any way to be "without a paycheck" for any length of time. Even if these folks ultimately want to go into academia, nothing is stopping them from landing a visiting scientist position at a company or something else that might end up being a very rewarding long-term career.

      If someone has gotten to a point in their career where the only thing they can do is academic-style research, that seems like a really bad place to be, and doubly bad for them to have chosen to not take a tenure-track position in the first place. What were they thinking?

      But I completely disagree that any of these people are in that situation. There are lots of very exciting jobs they can have, as long as they are open minded. We're talking about some of the smartest people in the field here -- certainly they are smart enough to figure that out.

      Delete
    5. Maybe not while employed at Twitter, but Google Research still exists, Facebook Research is starting up and so on. Do you think Omer Reingold will remain unemployed ? :) As far as months without employment - not a microsoft employee myself, but severance packages tend to be very generous for big companies.

      Delete
    6. Please tell me anonymous Oct 16 2:26 PM did not compare working on research at a Google or Facebook to working at McDonalds. Get some perspective, people!!!

      Delete
    7. Exactly. That's like suggesting that a physicist could overthrow centuries of belief in the Newtonian model of the universe while working as a patent clerk. How absurd.

      Delete
    8. Matt, you are drilling down on the word jobless trying to find some "absolute" meaning for it. When humans communicate, we always use words in context (it feels strange that I have to point this out to you, but it looks like I do).

      Someone is jobless given that they are trying to stay on their career path, someone else is jobless given that they are looking for a job that pays enough to support their family, etc.

      Delete
  5. The thread that runs through all these successful industrial labs is that they were supported by monopoly profits: AT&T by the phone system, IBM the mainframe, Microsoft Windows/Office. As each institution lost their profits, they scaled back their research. Microsoft will be no exception, the only question is how long it takes. I give them 10-15 years.

    ReplyDelete
  6. Anonymous Oct 16, 2:14 PM has it exactly right. In the case of AT&T (and arguably in the case of IBM), the monopoly involved an intricate set of checks and balances between the company and government where the company's research was seen a public benefit in exchange for monopoly rents (that's a reason why AT&T in pre-divesture was not allowed to benefit from patents outside its core business in telecommunications). However, times have changed, and sometimes not so uniformly against industrial research. I've been able to organize and grow long-lived research projects at Google way more easily than I was able to in academia (even though I had a well-funded academic research group overall). Of course that has much to do with the fact that my research interests are well aligned with Google's mission, but the overall climate for sustained research projects in academia is grim currently, so Matt's provocative take is worth considering. Finally, whatever the reasons for MSR-SVs closure, what I do know is that friends and acquaintances affected by it were really shocked by how it went down, and are struggling to figure out how to preserve the closely-coupled research partnerships they had built with so much effort. In other words, the losses are not so much of fungible "jobs" as of a subtle fabric of collaboration that will tear irreversibly as its nodes disperse among different institutions. In this, MSR-SV was very different from academia, where individual faculty members are sole proprietors who may occasionally collaborate (mainly through graduate students) on projects, but are basically independent from each other.

    ReplyDelete
    Replies
    1. Can you elaborate more about "I've been able to organize and grow long-lived research projects at Google way more easily than I was able to in academia" ?

      The thing person A means as long-lived project (say a self-driving car) and the thing that person B means as long-lived (say p=np?) work in totally different ways... I can't imagine an industrial lab hiring 5 researchers to just work on p=np for 10 years straight...but i can imagine an industrial lab doing the same for a self-driving car..

      Delete
    2. I work on AI: natural-language understanding and machine learning. The ultimate objectives very far away, but intermediate steps have been quite valuable at shorter time scales.

      Delete
  7. Hi Matt,

    First off, I think you provide a valuable perspective, one that many people are thinking but afraid to state publicly. However, I think there's one very big point you are missing, the motivation and reason behind the letter. I think many researchers were very shocked and upset by the handling of the SVC closure. The researchers were laid off immediately with very little notice (notice on Thursday and closure the following day). This might not seem like a big deal, but one very important thing to consider is the IP arraignment between Microsoft and the researchers. Another point to note is that the NSDI deadline was the following week. As a result, many researchers were not able to submit their research that they were working on for months because of IP issues. Moreover, all the research they have done at MSR is now in limbo from a legal perspective because they no longer have access to their corporate laptops. I think people in the academic community would have liked to see a more "gentle" phasing out (SVC will be shutting down in a couple of months. Please have your research in order and plan for this) rather than a sudden corporate-style layoff. This type of layoff doesn't even happen in academia when people fail to receive tenure.

    I think it's important to acknowledge this and understand that by doing this, Microsoft is treating MSR in a very corporate manner, which opens the door to many questions about the future of the organization.

    ReplyDelete
    Replies
    1. "Microsoft is treating MSR in a very corporate manner" - as an industry person, Microsoft treated the researchers at MSR-SV in the same manner they treated other MSFT employees. This type of layoff may not happen in academia but it's the norm in industry (regardless of the job description of those laid off). As someone who's gone through layoffs and as an industry researcher who is a huge fan of MSR, I understand the disruption caused by this sudden lab closing but again, this is the norm and it's part of the risk of coming to industry. Academia comes with a different set of challenges (obtaining grant money, etc.) - it's shocking to read of labs (especially in life sciences) closing down due to lack of funds, etc.
      Like many others, I bemoan the end of the large, academic-style labs (of which MSR may be the last one standing), but the changes in industry labs (closings, outsourcing research to India or China where many brilliant people can be hired for the price of few here) are real and grad students should be adequately prepared.

      Delete
    2. Yep, that stinks. I still think the letter is overly pessimistic about the implications for industrial research. The world can live without a few NSDI papers, and these people will continue to have great careers.

      Delete
  8. I have read you blog in the past and enjoyed many of your posts. This post, however, I must take offence with.

    I could point out that the Google research model maybe better for some types of research, but by the own admission of a panel at a Google event for SVC last week, Google is not the place to do certain other kinds of very important research (i.e. that does not ship something in a year). I could mention that the "generous severance and stocks" would not pay the rents until next summer for people who were at SVC "only" five years. I could mention that interviewing for academic jobs normally entails two months of traveling every or every other week and that may be somewhat inconsistent with what any bay area company would expect from a research scientist.

    But I think the main point I see you missing is the following. The main reason for this letter is that this recent lab closing marks a fundamental change in the implicit contract that people going to research labs believe they are getting into. And the implications of this change may be felt for a long time certainly by Microsoft and possibly by all of industry.

    What contract, you ask? Research labs have died numerous times in the past and it would be foolish to expect a research lab job to be like a tenured position, you'd say. Yes and yes. But when research labs have died in the past, it has, in every single case, either been because the company was severely hurting financially, or at the very least, a long slow shift in focus by the company away from research. In either case, a researcher at one of these labs had a two-, if not five-year period when the writing was clear, either in the company's financial reports, or in the dwindling travel budgets and the immediate-internal-impact-focused annual evaluations. This gave them sufficient time to adjust to the changed nature, or move to another possibly academic job. This is what anyone joining MSR or any research lab expected: it won't last forever but we would have a few years to move to a different position when our work is no longer appreciated by the company. Such an implicit contract has allowed over the years several excellent researchers, even the less brave ones, to try working in such labs and then either decide that they like more direct impact and move closer to industry, or move back to an academic job but with a better understanding of the real world and a resulting impact on their research interests. This has benefited industry as well as facilitated greater and faster exchange of ideas between industry and academia. This, I believe, has impacted positively the impact CS academic community has had on the world, including things like the founding of Google.

    Breaking this contract therefore will hurt both sides. This letter is, in my view, an attempt to convey the belief that this particular incident is a freak outlier, and not a harbinger of a fundamentally altered relationship between industry labs and academia. That is the message they are sending to students and junior researchers.

    ReplyDelete
    Replies
    1. I'm in complete agreement with you. If the letter had been worded in that way, and had expressed those sentiments as well as you do here, I wouldn't have seen reason to take issue with it. My concern is that the letter is overly shrill when it comes to the immediate, short-term implications of the lab shutting down for (some) of the affected researchers. Perhaps in trying to make a point about the broader implications for encouraging junior researchers to join industry, the authors of the letter just got carried away with claiming these people would be "jobless". I hope you agree with me that claim is somewhat ridiculous.

      Delete
    2. I could not attend the event last week because of a prior commitment, but I'm startled by "the own admission of a panel at a Google event for SVC last week, Google is not the place to do certain other kinds of very important research (i.e. that does not ship something in a year)." I've been at Google for almost 7 years, involved in multiple successful projects in machine learning and natural language processing. Most of those have taken much longer than one year to bear fruit.

      Delete
    3. Matt, if you agree with Anonymous @ 8:53am, then you agree with the original letter. It looks like the whole reason for your post is that you took the word "jobless" out of context, and then launched into a long and elaborate discussion that is at best tangent to the issue at hand.

      Delete
    4. Please go back and read my entire post. It's not just about the word jobless -- it's the fact that the letter is off base on many aspects of what the career prospects are for the folks let go by MSR SV.

      Delete
    5. Look, I read the entire blog post before I posted the original comment. My reading is that you disagreed about whether the researchers would be jobless. From that, you launched into a 1,600 word (!!!) blog post that is rehashing the same issue.

      Delete
    6. A more constructive way to think about this is that Anonymous @ 8:53am and the letter are saying pretty much the same thing (except for minor changes in wording). So if you agree Anonymous @ 8:53am, you agree with the original letter, except for wording issues.

      Delete
  9. Matt, congratulations on your thought-provoking post, and for generating such a thoughtful discussion. I especially agree with your point about the tone of the letter. And I also agree with your point that the world has changed and rather than clinging to bottoms-up, basic research models of the past, this group of notable academics should be focused on proposing new, more flexible models that can work in the future. Again, congrats on the post as you've generated a really thoughtful and interesting discussion about the role of basic research in our society.

    ReplyDelete
  10. SIGACT's reaction and some of the comments on this blog seem to see academics and researchers as a better breed which is not subject to the same laws as the rest of the workforce.

    I am certainly sympathetic to those affected by the closure of MSR SV. Many fine colleagues are now in search of new jobs. I also remain a strong supporter of the role of fundamental research in society, hoping our work continues to be funded, somewhere and somehow.

    Nonetheless, what shocks me the most is the implicit understanding that in a restructuring operation that resulted in a loss of jobs in excess of 10000, some think that seventy or so researchers should have been treated far better than the rest because of the higher intellectual value of their (non-revenue producing) occupation.

    It comes without saying that many among the non-MSR layoffs are also facing considerable life challenges, such as paying their mortgages or their student loans, dealing with child care, and supporting their ill relatives. Some of these people may not be sufficiently qualified to transition very quickly to a new opportunity. Some of these will do fine, other will be the main actors in tragic stories we will never hear.

    Compared to this, the fact that some former MSR members may not be able to submit their NSDI papers seems like a negligible loss. Implying that Microsoft shows little touch by not having considered this fact is incredibly naive.

    Fortunately, I am confident that our students will not be intimidated by the badmouthing promised in the letter. Perhaps showing more maturity than their teachers, they are completely aware that IT jobs are ephemeral, and that they should not base their life plans on the expectation of a long-term position.

    Microsoft's move was perhaps harsher than what we have seen in the past, but was also completely in line with what IT professionals are already used to see. It also does convey a clear message to the rest of MSR, giving to the majority of MSR folks still ample time to assess where the wind is turning, and to react accordingly.

    ReplyDelete
    Replies
    1. great comment. I completely agree.

      Delete
    2. Agree. I think this implicit hierarchy has been hurting Microsoft before. A CS graduate working at Microsoft engineering department was automatically considered inferior than his MSR counter part, even though he is the one who actually making money to support those non-revenue generating research.

      Delete
    3. Not better breed but different profession. And NSR-SV more than paid for itself.

      Delete
    4. Great comment. The letter seems to imply that Microsoft "owes" more to a certain type of employee than to others and furthermore that Microsoft "owes" an explanation and perhaps reparations of sorts to the academic community. I would note many people with varying job descriptions are recruited by corporations with pitches which may or may not turn out to be true depending on the business conditions - additionally, it is not the job of a company to care about one's particular career path. While the affected researchers are probably reassured by the support expressed in this letter, some of the language is hard to read. Industry can be harsh and bad things happen to good people all the time, as the comment above details.

      Delete
    5. If MS doesn’t owe any explanation to academia, then they should have no expectation of cooperation either. This is all the letter says.

      Delete
    6. There are many types of academia-industry cooperation (or collaboration). It may be the case that some of them are not well-served by current realities in industry and may be risky for academics. I think it's fair to state that and to act accordingly. However, unless all current Ph.D students are heading towards the tenure track, industry research labs - with the associated risks - will remain viable employment options and students will still benefit greatly from cooperation in the form of internships or joint projects of shorter duration.

      Delete
    7. So much is being made out of MSR's decision to close the SV lab. I don't recall there being a similar amount of fuss when Intel decided to shutter its lablets. Those lablets did great research, and collaborated closely with academia as well. I don't know what came out of them to benefit Intel's bottom-line, but they resulted in advances in many fields of CS. Why the noise now when MSR decides to do this, but not the same level of noise when Intel (also a very rich company) decides to do the same?

      Delete
    8. The letter and surrounding discussion are likely due to the exalted position of MSR, which was in fact one of the remaining academic-style labs. There is a "et tu, Brute" attitude on display, as MSR layoffs confirm wholesale changes in the industry research lab model. These changes have been going on for a while, but not sure academics had noticed.

      Delete
    9. The point is not to treat MSR better than the rest of MS. Industry employees have it better than academics in some ways, and academics have it better in other ways.

      The point is that since these people are academic researchers, they have to go on the academic job market when they are out of a job. In the past, Microsoft has represented that it understood that, but now it looks like it doesn't.

      Let me give you an example. Jim is an intense and successful high school football coach. One day, Jim start offering his services as a high school baseball coach. After his intense workouts, his best pitchers are sidelined for months with serious elbow and shoulder injuries that threaten their entire careers.

      When people confront Jim about this, he sidesteps the main issue and angrily claims that "baseball players want to be treated better than football players".

      Delete
    10. An MSR employee IS an industry employee. While MSR may have represented itself as an academic-style lab, it is not a university, but part of of a company. For many years, MSR employees were in fact treated differently and Microsoft encouraged top academic researchers to join MSR by promising an academic life without the hardship of grant applications or the distraction of teaching. However, the contract states that you are an "at-will" employee, not a tenured faculty member. I understand people feel betrayed, but MSR/MS is the industry.

      Delete
    11. Look at the comment above, man. Baseball players are different from football players. I think that explains it best.

      Delete
    12. Am not arguing what MSFT did was right, or fair, or not impactful as far as folks' careers are concerned. I fully understand the difference between a researcher and an engineer but I also understand the difference between a university researcher and a industry researcher (even if the industry researcher works at MSR or other seemingly rock-solid places such as Google Research).

      Delete
  11. I think you are missing the point of the letter. The way I read it, it's not so much about the individual researchers. I am sure that many of them will actually be jobless until July or September, since they would rather be a few months without a salary than move away from academic research. But I agree that on the grand scheme of things there are worse tragedies than some people that had very nice salaries until a month ago being few months without pay.

    The point is that MSR has a particular model of industrial lab research which is attractive to certain kind of people that place a high value on the ability to set their own agenda and publish their work. You may think that this is not the right model for industry research, and such people should have never taken an industry job in the first place but rather a tenure-track jobs. But at least until last month, and arguably still today, Microsoft and MSR management had a different opinion and MSR has actively tried to recruit such people. Indeed some of those people took MSR jobs over tenured positions and others preferred MSR to much higher offers at other companies precisely because of this reason. (And sometimes these people that would have never come to MSR if its culture was different, ended up making very significant contributions to Microsoft's business.)

    Now, if MSR had decided to move away from that model, then professors should in fact not recommend it as highly as they have done before to the kind of people that value these types of academic jobs. But if MSR did not move away from that model and they still wants to recruit the people who would otherwise be in tenured or tenure-track positions, then they would be wise to listen to this letter.

    ReplyDelete
    Replies
    1. I see no evidence that Microsoft is moving away from its academic research model. They were no doubt forced to cut back amidst a large round of layoffs across the company. At least closing a whole office avoids singling out any individual researchers, so anyone let go can chalk it up to corporate restructuring rather than any negative reflection on their own performance. In a way, it was the smart move.

      Delete
    2. You may not know MSR that well then. This was the largest unit that was run in an academic spirit, and where impact on the scientific community was valued as much as impact on products (of course, impact on products was naturally better rewarded). Redmond has tiny islands of this kind and NE has a very small head count of such people. I think that people within MSR understand the message and reacting appropriately.

      Delete
    3. By letting go of academic researchers like this, Microsoft has pretty much abandoned the academic research model. The point of the letter is to talk with Microsoft to see if they really meant it, or maybe they just didn't appreciate what they were doing.

      The immediate impact is that the researchers who are laid off are likely to be jobless for about a year. The larger impact is that they will all go on the academic job market *at the same time*.

      Moreover, every other academic researcher at MSR now sees the writing on the wall, and will most likely go on the academic job market, also at the same time. This includes people who left tenure-track and tenured jobs to join MSR.

      The net result is that many excellent researchers who would be on the tenure-track or tenured if MSR simply did not exist are now in a tight spot and could easily end up with their academic career derailed over the long term.

      Delete
    4. I'm sorry, but I simply disagree with both the assertion that Microsoft has "abandoned" the academic research model. From what I can tell they still employ some 900+ "academic" researchers. The fact that they had to cut back by a small percentage does not represent abandonment. I don't understand where that comes from.

      I also simply fail to understand why so many people on this thread (and the authors of this letter, apparently) think that the only jobs for these folks are academic jobs. Just like lots of former professors (like me) have moved into industry, most of these people are very well-suited to other jobs in industry. There is, in fact, a world beyond academic-style research.

      Delete
    5. 1st paragraph - My impression is that most researchers at MSR (even the former MSR SV) are industry researchers. If they did a good job working on robotics or machine learning at MSR, they can probably get a robotics or machine learning job at Google, etc.

      It is mostly mathematicians / theoretical computer scientists (and maybe they have other kinds like theoretical physicists, I don't know) that are the issue in this letter. It is no coincidence that pretty much all signatories of the letter are TCS people, even though TCS people were a minority at MSR SV.

      Unless Microsoft comes to some agreement that this is not how it is going to run its labs in the future, MSR will be an infeasible place for people who want to build careers in TCS. This is not a demand, more like a physical reality. People at MSR who can leave will leave, and new people will not go to MSR except as a last resort.

      2nd paragraph - The first thing is that you *wanted* to leave. The second thing is that you were a professor in CS systems, not theory. CS systems to software engineering is probably the closest transition one can make. You can also make a plausible argument that you can have a bigger impact in CS systems as a manager at Google than as a professor.

      This is decidedly not true about theory. So these people are faced with losing their profession because they joined MSR, even though back when they joined they had tenure or tenure-track options.

      Delete
  12. "The authors of this letter seem to believe that doing academic-style research is some kind of special, protected occupation that people with PhDs are entitled to."
    It was never stated in the article that only people should with PhDs should do fundamental research. By all means, a high school kid is free to sit at home and try to think about the origins of our universe or the Riemann hypothesis, nobody is stopping them.

    "While there may have been plenty of industry labs doing pure, academic-style research 20 or 30 years ago, the world has changed. What worries me is that professors still cling to an industry research model that has largely been supplanted by other, better models."

    Yeah the current models are far better (at making money), I agree. Maybe they should just make all the researchers work on the next 'snapchat' idea and make even more money instead of all this fundamental research crap.

    "It is also a common misconception that one cannot do research in a more product-oriented industry setting. For example, Google does a tremendous amount of research, albeit using a fairly different model than places like MSR. The same is true for other large and small companies as well. We may not call all of these places "research labs" and bestow fancy titles on people like "research scientist", but the research is still happening, and having tremendous impact."

    OMG, seriously! This is so stupid, that I do not know what to say. You are comparing two totally different things. Yeah, who cares what kind of research it is, as long it is research. Surely, all the theory people should stop thinking about P vs NP, instead they should join facebook and do research on how to make farmville even better, it too is research, surely it will have tremendous impact!
    --
    People working in fundamental research will be far more beneficial to this world over long term. Most of the research done in the industry research labs is very narrow and focused on making products. It is in the interest of the world as a whole, to respect people doing fundamental research, because ultimately these are the most precious things that we pass on to the future. By respecting them and nurturing them, we can encourage more people to work on fundamental research. Most of the "tremendous research" happening in the tech companies that you refer to, will be totally useless in a few decades. For example, all the light manufacturing companies in the early 20th century did tremendous research to make the light bulbs maybe 10% more bright, or 10% more durable. But now we have much better LED lights, thanks to some crazy physicists trying to make some esoteric quantum theories at that time. Surely, they should have instead done the "tremendous research" at the light bulb companies to make their world a better place.

    All these companies now, which are making tremendous profits based on the work of thousands of brilliant scientists and mathematicians, should try to contribute some of their income in funding fundamental research. You can call it charity, but it is the charity of all those brilliant scientists that these companies are living on.

    If you are not capable to or not interested in expanding the frontiers of sciences and mathematics, at least respect and help who are doing these things. Just don't bash on them.

    ReplyDelete
    Replies
    1. If you actually believe that Google is working on research which will be "useless in a few decades" and that Facebook's research arm will focus on improving FarmVille, you do not have a clear picture of the scope of research in the much-maligned industry research labs.

      Delete
  13. The vast majority of research candidates don't make it through to faculty jobs, and waste prime formative years of their career pursuing them, all the while underpaid and subject to major volatility and terrible work life balance. The academy does little to obviate this issue. Why should they? This glut of candidates allows them to be extremely selective. It's exploitation.

    Industry has layoffs, sure. But the average experience is much much better in terms of pay, work life balance, and volatility.

    It's quite hypocritical (or, at least, extremely self-unaware) of the academy to posture as though they are offering something comparatively valuable to the candidates from which their field is built.

    ReplyDelete
  14. And MSR comments back: http://blogs.msdn.com/b/msr_er/archive/2014/10/21/harry-shum-open-letter-to-academic-research-community.aspx

    ReplyDelete
    Replies
    1. A very corporate response (not that anything else was expected). I work in industry and would caution any student who is interested in an academic career in the long-run to come to industry with their eyes open. You can get enormous benefit out of being in industry but please be prepared to be laid off even if the company can afford you.

      Delete

Startup Life: Three Months In

I've posted a story to Medium on what it's been like to work at a startup, after years at Google. Check it out here.