Tuesday, November 16, 2010

Guest Post: Why I'm staying at Harvard (by Michael Mitzenmacher)

[Michael Mitzenmacher is a professor of Computer Science and the Area Dean for Computer Science at Harvard. He is a dear friend and colleague and has been one of the role models for my own career. Michael wanted to respond to my earlier blog post on leaving Harvard with his own reasons for staying; I am only too happy to oblige. (I swear I did not ghost write this.) You can read more of Michael's own blog here, though he's not posting much these days. --MDW]

To begin, I'd like to say how sorry we are at Harvard that Matt's not returning.  Matt's been a great colleague, continually pushing to make CS at Harvard better.  His enthusiasm and tenaciousness have made us tangibly better in numerous ways.  I, personally, will miss him a lot.  Matt pushes hard for what he believes in, but in my experience he's always done so with open ears and an open mind.  We're losing a leader, and Google is lucky to have him.  I have no doubt he'll do great things for the company, and maybe even earn them another billion or two.

While Matt's decision has been a blow to CS at Harvard, I'm optimistic that our plan for growth will, eventually, make up for that loss.  My job as Area Dean is to try to make that happen as soon as possible.  I don't want to suggest that replacing Matt will be easy, but rest assured we'll be on the case.

I'd also like to say that I think I understand Matt's reasons for leaving.  I'm glad to have him write "I love Harvard, and will miss it a lot."  And how could I disagree with statements like "The computer science faculty are absolutely top-notch, and the students are the best a professor could ever hope to work with. It is a fantastic environment, very supportive, and full of great people."  But I know from previous talks with him that he hasn't always loved being a professor.  And that's what I'll try to write about the rest of the post.

I think there's a sense in academia that people get PhD's so that they can become professors.  Most graduate students have that point of view going in -- their experience with research professionals at that point is essentially entirely with faculty.  And most professors encourage students to have that goal.  Some of that, I think, is that most professors like their job (unsurprisingly), and some may not have other experiences to suggest to their students.  And some of it may be more calculated.  One measure of a faculty member's success is how many faculty offspring they've produced.

But being a faculty member is not for everyone.  As Matt has described in this blog, and I in the past have described in my blog, being a professor is probably not exactly what most people expect.  Besides teaching and research, your time gets taken up with administration, managing (graduate) students, fundraising, and service to your scientific community.  It's perhaps absurd to expect that everyone who starts out in a PhD program be interested in all these various aspects of the job.  And, fortunately, in computer science, there are still many other compelling options available.

As Matt says, at Google, "I get to hack all day."  That's just not true as a faculty member -- time for actual hacking is usually pretty small, and more of your time is spend managing others to hack for you.  (This is a complaint I've heard from many faculty members.)  I can understand why Google would be a very appealing place for someone who wants to write code.  I'm sure Matt will come to miss some of the other aspects of being a professor at some point, and I'd imagine Google will to some extent let him entertain some of those aspects.

One of the comments suggested money must be a motivation.  For some people who have to make this choice, maybe it is.  (See Matt's comments on the post below for his take on that.)  So what?  Again, it's good that in our field there are good options that pay well.  That's a big plus for our field, especially if we accept the fact that not everyone can be or wants to be a professor.  But as Matt says, professors at Harvard (and top 20 institutions in general) are doing just fine, and money probably isn't the main issue for those who choose a different path.

I suppose the question that's left is why I'm staying at Harvard -- that is, why I still like being a professor.  (And thank you to those of you who think the obvious answer is, "Who else would hire you?")  I enjoy the freedom of working on whatever I find interesting; being unrestricted in who I choose to talk to about research problems and ideas; having the opportunity to work with a whole variety of interesting and smart people, from undergraduates to graduate students to CS colleagues all over the globe to math and biology professors a few buildings down; the ample opportunity to do consulting work that both pays well and challenges me in different ways; the schedule that lets me walk my kids to school most every day and be home for dinner most every night; and the security that, as long as I keep enjoying it, I can keep doing this job for the next 30+ years.

The job is never boring.  On any given day, I might be teaching, planning a class, working with students, thinking, writing a paper, writing some code, reading, listening to a talk, planning or giving a talk, organizing an event, consulting in some form, or any other manner of things. In the old days, I wrote a blog.  These days, I'm administrating, making sure our classes work smoothly, our faculty are satisfied and enabled to do the great things they do, and we're able to continue to expand and get even better.  Once I wrote a book, and someday I hope to do that again.  Perhaps the biggest possible complaint is that there's always something to do, so you have to learn to manage your time, say no, and make good decisions about what to do every day.  As someone who hates being bored, this is generally a good feature of the job for me.

And Harvard, I find, is an especially great place to work.  We attract some of the most amazing students.  Our still small-ish CS faculty really works together well; we all know who each other are, we keep aware of what we're all doing research-wise, we collaborate frequently, and we compromise and reach consensus on key issues.  Outside of the CS faculty, there's all sorts of interesting people and opportunities on campus and nearby.  Boston is a great city (albeit too cold and snowy in the winter).

Other profs have made similar comments in Matt's post -- there's a lot to like about the job, and at the same time, it's not the best choice for everyone.  Of course I don't like everything about the job.  Getting funding is a painful exercise, having papers rejected is frustrating and unpleasant, and not every student is a wondrous joy to work with.  I sometimes struggle to put work away and enjoy the rest of my life -- not because of external pressure (especially post-tenure), but because lots of my work is engaging and fun.  Of course that's the point -- there's good and bad in all of it, and people's preferences are, naturally, vastly different.  I don't think anyone should read too much into Matt's going to Google about the global state of Computer Science, or Professordom, or Harvard, or Google.  One guy found a job he likes better than the one he had.  It happens all the time, even in academia.  It's happened before and will happen again.

But I'm happy with my job right now.  In fact, I'm pretty sure my worst day on the job this year was the day Matt told me he wasn't coming back.  We'll miss you, Matt, and best of luck in all your endeavors.


  1. Mike is theory and Matt is systems.

    Although there are theorists doing pure theory in industry (mostly at Microsoft it seems), I think it is easy to see why academia might be especially congenial to a theorist. Even a theorist who is close to systems people like Mike.

  2. I'm in systems just down the road from Matt and Mike. I don't think I could have better said why academia appeals to me than Mike. Fundamentally, it's just very hard to find a job with more flexibility and freedom to do what you want than academia. I'm my own boss -- besides 3 hours of teaching a week (9 months out of the year, and 7 years out of 6!) my schedule is completely under my own control.

    I will add that I thing one thing neither Matt nor Mike pointed out is the opportunity that academia provides for commercializing your own technology. Most universities are extremely supportive of their faculty starting companies, and these days it is very common for faculty at universities in startup-friendly towns to spend a significant portion of their time at a startup. Matt complains about not being able to have impact or get your ideas accepted in a company. What better way to do it than to found the next VMWare or Akamai? Sure, you can go to Google, and make money for someone else and work on their vision for the future, or you can make your own way! Starting a company as a grad student is certainly possible as well (witness Google, Yahoo, etc) but having a tenured job to return to when your startup doesn't quite work out as planned sure is nice...

  3. Sam -- great point. Harvard does not have much of a startup culture amongst the CS faculty. I'm also too lazy to do a startup myself -- again, it comes down to overhead (selling something) vs. hacking.

  4. @anon1: I'll let others discuss a potential point Matt raised that you seem to be echoing -- whether "systems research" and/or "systems work" is better done at companies that universities. I personally think there's plenty of room for such work at both, and that again it's a matter of one's own desired working style.

    @Sam: Good point on starting one's own company. I haven't done that, but have been happy consulting for startups when that's worked out.

    @Matt: I think you're exaggerating Harvard's "lack of" startup culture a bit there. Offhand, I know Mike Smith, Margo Seltzer, and Stuart Shieber all started successful companies. It happens here too.

  5. What seems to be missing from the discussion is the fact that professors, who are typically hired because of their research and development skills (published papers), are forced by the organization and incentive structure to become cheerleader managers. Does this not seem like a bad idea? The org chart of academia is flawed in the sense that the top researchers are wasting their time not doing research. Instead they have to spend a majority of their time working on grant proposals, reports to funding agencies, administration and teaching. These positions should be filled by people who train themselves as specialists and feel passionate about their role, not top researchers who would rather be in the lab working with grad students (kind of sounds like Google). Some professors manage to figure out the teaching component, but others were never cut out for it. No wonder, since most profs have zero teaching skills when they start, and in most institutions tenure is a result of publications, not happily educated kids.

    Besides theory people, who basically have nowhere else to get paid for their work, it seems like becoming a professor is one of the worst decisions you could make as someone who actually wants to DO research.

  6. I feel the need to chime in on the lack or not of a startup culture at Harvard -- it always amuses me when people say that Harvard doesn't have a startup culture. I believe that if you count number of actual companies started per faculty member, Harvard CS has got to have about as high a ratio as anyone (except perhaps the institution Stonebraker resides in, since his rate of startups per year astonishes all).

    Off the top of my head, with our current 17-ish people, I can count at least five companies, and that's without thinking very hard. While a small number, it is 25% relative to our faculty, and that's a pretty big percentage. (I'm ignoring the companies started by our grad students and undergrads, which I'll also stake against any other institution's.)

    So, why do we get dismissed on this front? Rather than speculate, I'm sure I'll start a fire storm here ;-)

  7. If I may play the advocate's devil...

    What is the impact of the startups of Harvard CS (compared to the ones for instance started by Stanford CS)?

  8. I just have to comment on this one because my own experience is a hybrid of the alternatives discussed above and of course, that there are several strange coincidences too boot.

    Ten years ago I was teaching at Harvard (Finance in the Business School) but wanted to hack more (just like Matt), because I worked in Financial Engineering, which involved both high-performance computing and algorithms theory. With no CS background at all, I took a sabbatical to go to Berkeley (CS Theory) to tool up, hack, and generally indulge myself, and coincidentally started there with Matt as graduate student (if you remember) and worked with Alistair Sinclair (who is Mike's advisor I believe!). While in the Bay area, I was offered a job at SCU where they promised me I could work in both Finance and CS and they would value both (believe me, every other school in the Bay area would not value work across fields). Which I continue to do till today. I never went back to HBS, even though I enjoyed every moment of my time there.

    Both Matt and Mike are right, there are great things about being a professor and wonderful things about being in full-hack mode. I think I managed to get a bit of both and that works too. There seems to be a working assumption that once you leave academia, you never come back, or that no one will want you back. For someone talented and creative, especially in CS, this is hardly true. Most of us academics remain where we are because of the freedom and job security, but it also makes us lazy, and if we seriously think about it, twenty years into the game, no matter how productive we are, we might do greater things with a little pressure and focus that is more than just peer-generated. There is also something magical about the Bay area and the firms there that make it easier to leave. But more generally speaking, there is a lot of movement now between academia and industry, especially in the two fields I work in, and both are much better for it.

    So good luck Matt, from "Harvard to Hacking" sounds like a terrific idea to me!

  9. I guess I stand corrected on the startup front. My reference point is Berkeley and Stanford which is kind of hard to make comparisons with. I think most of those companies were started in the "good old days" before the Internet bubble collapsed, before I got to Harvard. It's also true that since I've been a faculty member there, there has not been that much VC and startup activity *in general* in the Boston area - and even if there was I was probably too busy trying to get tenure to notice :-)

  10. @anonymous: well, by Stanford startups, I assume you mean Google. I'll see your Google with a Facebook and raise you a Microsoft (since Google was a student start up not a faculty startup anyway).

  11. Matt -- for visible startups in the Boston area (not necessarily related to Harvard), how about Akamai, ITA software, Endeca, etc.

    Yes, they're not startups now, but they're companies that lasted through the bubble. They're the ones we know because they're still around and big. There are plenty of other startups around you haven't heard of -- both because they're not that big (yet), or not in your area, or (as you say) you've been a little too busy with that tenure thing to notice.

  12. @Jeff: I expected and predicted the "Mike's a theorist, and they can't get jobs in industry, so of course they want to be professor" type response, and even joked about it in the text. But seriously, you're mistaken.

    If you want to do research, there are plenty of companies arounds with research labs. Just like in systems.

    And if you want an industrial-style job, you can get that too. Yes, you'll probably have to code, and not have the chance to do real theory-style research-work, as you'll be building systems. But that's the same for systems people -- most systems people (even at Google) don't do real systems-style research-work, but spend their time building real systems, which may not be research-interesting. I expect the number of engineers at Google doing research-style systems work is relatively small, and even as a comparison point, Google at this point has hired a number of theorists (and statisticians, and machine learning people, and...) that do research-style work in their area too.

    This canard that theory people "basically have nowhere else to get paid for their work" isn't particularly helpful. It's roughly as bad as saying that Harvard/Boston doesn't have a startup culture.

  13. What advice do you give to your PhD students who are not hired by a Top 20 CS dept? The struggle to get funding is worse for new faculty at lower ranked depts. There's also the opportunity cost of working 5 years and not getting tenure. Google et. al. are competing for talent with the < top 20 depts.

  14. Michael - I already let you have the guest spot on my blog, so be nice :-)

    My point is that compared to the Bay Area, Boston startup culture is pretty small. Nobody would argue with this. Don't forget I was at Berkeley at the height of the boom so I have a particularly skewed perspective. One metric is the number of CS faculty taking leaves to start companies. At Berkeley and Stanford in the early 2000's there was a tremendous amount of that happening. Since I've been at Harvard I don't think it has happened at all, unless I'm forgetting someone.

    I don't agree that Google's systems work is not research-interesting. Google publishes a huge number of papers across many disciplines - see: http://research.google.com/pubs/papers_by_year.html

    You are right that the focus is on building and not doing research for research's sake - though there are a number of pretty out-there "researchy" projects at Google, like the self-driving cars.

  15. @Jeff, when you say "professors [...] are forced by the organization and incentive structure to become cheerleader managers", that's just wrong. It depends very much on the professor's style. That's something for students to think about when choosing an advisor, since it can affect the environment substantially.

    In fact it's easier to do a time-varying mix of management and hands-on work in academia than in industry, in my experience (campus, industrial labs, startups.)

    FWIW, I personally don't find I waste that much time on grant-writing. I think that's a red herring. Mentoring grad students is the largest commitment, and that's adjustable over time.

    To me, the most significant comment/question in this thread is whether the general discussion is relevant outside the top n schools for some modest value n. Probably not; I think most of the faculty posting here (myself included) are assuming the discussion is contextualized above that threshold. That was Matt's situation.

  16. @Margo: the devil's advocate here... (I was tired when I wrote my previous comment)

    Thank you for your witted comment :)!

  17. The choice between "hacking" and academic life is not necessarily an either-or. Industry research labs offer a hybrid, which in the best case results in a "best of both worlds" scenario: substantial freedom, low overhead keeping you away from actually doing research, paths to impact.

    (I left a top-tier CS tenure-track position after getting my 3-year reappointment, to take a research position in industry, for reasons similar to Matt's.)

  18. Someone from StanfordNovember 20, 2010 at 8:26 AM

    I'm not arguing that Google isn't a great place to work. But in academia you have the freedom to think about creating the future Google, and that's what academics should be doing. If you're doing work that solves Google's (i.e. today's) problems, yeah, its obvious you should be at Google and have direct impact.

    All the best!

  19. But in academia you have the freedom to think about creating the future Google, and that's what academics should be doing. If you're doing work that solves Google's (i.e. today's) problems, yeah, its obvious you should be at Google and have direct impact.

    No offense but that's the funniest thing I heard today. Because

    (1) Google, or Facebook for that matter, aren't started by university faculties. Oh wait, you're being sarcastic aren't you. Aha...
    (2) The academic world is pretty well separated from the 'normal' world. In the 'normal' world, you have to deal with folks of normal intelligence and below. You have to deal with reality, like politics and managing people with below average intelligence. Just by being an academic alone, you don't have that sort of experience. And whether an idea succeeds or not, depends a lot more on other factors, rather than just whether the idea wins the best paper award or not.

  20. Anon: "...deal with folks of normal intelligence..." -- what a narrow minded and arrogant comment. Wherever you are, this true. Plus in academia, you have to deal with people who will never survive in the real world due to social awkwardness. Think working with Sheldon from the Big Bang Theory. Yuck!

  21. Plus in academia, you have to deal with people who will never survive in the real world due to social awkwardness. Think working with Sheldon from the Big Bang Theory. Yuck!

    I agree partially :) I think in both cases there's a variation of talent, so in that sense you're right. The spread of talent in the industry, compared to say grad students at MIT/Stanford/Berkeley/CMU, is greater. Also, the average capabilities of the industry blokes and grad students differ significantly (otherwise the admission committees of those universities are going to be upset with you :D) And there's quite a number of folks in the industry who are socially awkward as well :)

  22. I have worked in both academic and industrial labs, and I think the elephant in the room is funding. First, US government funding is broken and rewards the wrong behavior. Search for any "center" and you will find a well-funded narcissistic weasel doing crappy work, and yet real people like the blog author have trouble getting money. Second, industrial funding is lacking. Why is there a a guy worth hiring for net $350k a year when he is not worth funding in a University with a group for the same amount. I assume there is a structural problem there as well, but I don't know exactly what it is.

  23. The most important computer software company in the world, Microsoft, came from Harvard, not from Stanford, Berkeley or MIT. End of discussion. All the top CS programs have a "start up" culture as well as an "academic" culture and a "relationship with large company" culture. This is a great thing.

  24. Yo,

    Whassup homies. This is heavy stuff, dawg.

    I dig the dirt that Sam Madden and MikeM kicked up. Its deep people.

    Ya'll forgot one more thing - Academics is also about teaching young minds to learn to learn. It is also about teaching young minds to love to learn. These skills are more important than ever.
    We've all had college professors/school teachers whose enthusiasm in teaching a class resulted in us students taking greater interest in the subject.

    The pursuit of knowledge and the synthesis of new knowledge is not a trivial sell to much of the people in our planet. Your job as a college Professor is an enormously important one in making sure that mankind moves towards a state where more of our decisions are based on knowledge and less on irrational artifacts like religion, blind faith, superstition, lies or historical notions that are not true. I would also argue that the pursuit of knowledge is also a very healthy way to live your life (as against using existing knowledge to shrink wrap ideas and make money).

    Let us not trivialize the magnitude of the contributions that Professors make to our Society to merely the set of tasks performed by a Professor during the period of a work week or a day.

    Peace Homies. Be Cool.

    gangstarapperla ATTTT gmail DOTTTTT com

    PS: I am not a Professor. I work in the industry. I am not a rapper - but I do like HipHop music.

    PPS: Matt - I live in Portland, OR and have many friends in Seattle (some who work at Google, Kirkland). If you need any help, shoot me an email, dawg. Seattle is "warmer" than Boston and you dont have those damn toll booths


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.