There's been an article going around about why not to get a Ph.D. (h/t John Regehr for the link). I'm using it as an excuse to toss out a bit of unsolicited advice about the Ph.D., and some anecdotes that perhaps some fresh entrants may find comforting. Or not.
For context, I loved my Ph.D. experience (after the first semester of hell when I was convinced I'd get kicked out for being too stupid. I'll return to this in a few days.). But I also have a very smart, competent friend who had an absolutely miserable, manipulative-evil-advisor kind of experience in a different program, and, while completing the Ph.D., ended up leaving the field entirely as a result of it. Both of us were at top programs in our respective areas. I was fortunate enough to have an incredible advisor and be in a program that generally supported its grad students pretty well. It matters. I'm also a ridiculous optimist, which I think helps a lot in research. (It doesn't make your research better, it just gives you an excuse to get through the down parts. :)
These aren't isolated incidents; you'll find plenty examples of people who loved the Ph.D., and plenty of anecdotes about hellish experiences, such as the well-known Corey lab suicide at Harvard Chemistry in '98.
What I'd like to say about this is, briefly: (a) I think that the post's author has some good points that anyone entering the Ph.D. should understand; (b) What I read from this is more a story about high variance and risk, as opposed to a universally negative experience; (c) Even people, like me, who loved their Ph.D. experience and emerged successfully are likely to have some down times, and there are some important things to keep in mind to get through them a better person, instead of becoming, as the OP notes, "broken."
High variance in quality of experience: The quality of your grad school experience is very strongly influenced by your relationship with your advisor and your peers. There's also a decent aspect of luck involved in, e.g., how much and when you publish. There's a lot of pressure, both internal and external. And this is all magnified by the fact that it's an extremely demanding pursuit: liv is correct; grad school is not a 40 hour per week job. It's more like 60, and in grad school, I ended up sleeping in my office more than a few times for major paper deadlines. A good grad school experience is great, in part because of that intensity, and a horrible grad student experience is pretty close to rock bottom for the same reasons.
Some of these factors are under your control: It's important to learn how to "manage up" your advisor. Faculty are, for the most part, insanely over-subscribed. Make it easy for them to help you: Send concise, easy to read emails that are easy to answer in short sentences. Seriously - this is under-appreciated in both academia and the real world. The ten paragraph-before-a-question email is likely to get thrown on the "answer soon" pile... which may or may not be touched that week. Almost every professor I know responds well to students who're self-motivated and organized. I _love_ it when my students walk into my office to discuss some aspect of research they're curious about or stuck on. That's my job - and it's one of the big reasons I picked this job. On the flip side, it's no fun to have a meeting a week later and say "um, didn't we already discuss this exact thing last week, and you didn't change any of it or write down what we discussed?"
Grad school (and academia) does expose you to a large amount of critique. Not all of it is constructive. This may come from your peers at your institution, your advisor, or (commonly) paper reviews. I'm not sure what to say here except that it's important to remember that you are not your research or your paper; you cannot control others' poor behavior, but you can control your reaction to it; and that even a shitty paper review usually contains information that you can use to improve the quality of your research or your writeup, and when someone hands you a pile of lemon-flavored poop, you can still use it to your advantage. And trust me: It's not just you. I've had a pretty successful career thus far, and some of the reviews I've received, both as a student and as a faculty member, would singe your eyebrows. I plan on a followup post where I show some paper reviews...
Others aren't: There are bad advisors, evil advisors, and bad student-advisor matchings. It can be very hard to differentiate these from a student not doing well enough or being up for the task. This is partly where your peers come in, even before you enter grad school: Advisors typically develop reputations. You can and should ask _other students_ what a professor is like to work for before you start working with them. But it's up to you, mostly, to decide if you are a good match with that particular style. My students would probably tell you that I'd be a bad advisor to someone who needed to be very closely managed and given concrete tasks to complete on a day-to-day basis. They'd be right. I'd go nuts and the student would go nuts. And that's _not_ because such a student is a bad student: Everyone starts the Ph.D. at a different point in their capabilities as a computer scientist and researcher. And every advisor has strengths and weaknesses.
Finally, mental health issues and the Ph.D. are an under-appreciated issue. Graduate students are a higher-risk population for mental health issues [Berkeley]. This is a tough one - it's simultaneously an individual, cultural, and institutional issue. I'll talk about this a bit below.
The OP discussed some financial issues that, for the most part, aren't relevant to CS Ph.D. students at good programs, where, for the most part, you're de-facto (or explicitly) guaranteed funding. Thank goodness. The Ph.D. doesn't pay well (in the two thousand+ per month range), but it's enough to live on, and most students also supplement their income with one or two summer internships that pay in the 6-7k/month range.
Not all of the things that need to be improved are under the control of the student. A few institutional ones include:
Spreading the power and responsibility: One thing I love about CMU is the degree to which they try to involve other faculty in the Ph.D. education process. Our "black friday" review of the Ph.D. students, for example, is more than 50% about creating an opportunity for other faculty to say "woah, are you sure you're doing the right thing by that student?". And, while I find it annoying as hell sometimes to have to force my students to jump through the hoops, I feel the same way about our department's recent emphasis on getting a thesis committee together early. It helps spread the responsibility of advising and reduces the risk and variance for the student. Near the time I graduated, MIT began a review process also designed to help students not fall through the cracks. There's probably more that can be done, but I recommend CMU's model as a starting point.
Better training and awareness of mental health issues: As a faculty advisor, I have received absolutely no training in how and when to steer students towards our mental health services folks. As a student, I received very little information about the availability and "ok-ness" of the need for such services. This is poor form and needs to be improved across the board. The Berkeley study I cited earlier has some good suggestions.
More honest support of non-academic post-Ph.D. career paths: As a chronicle.com article discussed several years ago, there's a pretty strong communication of the idea that the only true "success" after completing the Ph.D. is to go on to a top tier academic position. This pressure is real, and it's so bogusly stupid it's hard to know where to start. Particularly these days, the number of highly desirable academic positions is smaller than the number of good Ph.D. graduates. Not everyone wants to teach - Matt Welsh has some nice blog posts discussing some of these issues. It's very possible to have substantially more impact on the world by going off and creating a little startup, or completely re-doing Google's indexing system, and a lot of other things. This isn't a question of absolute value, it's a question of the academic culture rewarding students who make the same decisions as the people who ... decided to go into academia. And it's total shit - we need a culture change in academic computer science that values our students making the best decisions for themselves, not making the decisions that we want them to. I don't know how to solve this; I see a lot of students from CMU's parallel data lab deciding to go into industry, because they're exposed to a lot of positive examples from their interaction with industry. I think Stanford has an equally accepting climate of going off to a startup. But how to go from singleton examples to a field-wide acceptance is a good question.
All that said, I still believe that the problems with the Ph.D. process in CS result in high variance, not uniformly bad experiences. I, and many people I know, count their years in grad school as some of the best of their lives. Best advice is to go read Matt's advice about whether or not to get a Ph.D., and decide for yourself based upon whether or not your long term life goals are helped or hindered by doing so. Once you've made that decision, you can decide whether the specific Ph.D. opportunities available to you are likely to result in a satisfactory experience.
Popular posts from this blog
By David Andersen – February 12, 2013It's been a year and a half since I migrated my undergrad distributed systems course (15-440) to the Go programming language . I'm still glad I did it and will keep it there, but we've also learned some important lessons about how to best use it in our context that I hope I'll be able to improve upon next semester. Hopefully, some of this may be useful for those of you considering moving to or creating a Go-based class. The big points I hope to get across are: (1) Go is really good for distributed systems; (2) Students need more intro to the language than I thought; (3) The language has matured a lot, but there are some technical glitches remaining that you'll have to work around. Go is good for teaching (and building) distributed systems I believe this for two reasons: Go provides a good mix of high-level and low-level functionality. Channels and goroutines provide a program structure that works well for distributed programming. The importance
By David Andersen – December 11, 2013Over the past three weeks, I've been running crypto-coin mining software on between 20 and 60 Amazon EC2 GPU instances at a (very small) profit, with the goal of understanding the currencies, exchanges, and algorithms in more detail. This is post #1 of 2 discussing scrypt coin mining on Amazon. In the follow-up post, I'll go into the algorithmic and engineering details of building a better CUDA-based miner. A few weeks ago, Gün and colleagues wrote a paper describing a new collusion attack against Bitcoin . I hadn't thought much about BTC before that, but he got me curious. As I explored, I discovered Litecoin , an alternative to Bitcoin that is designed to reduce the comparative advantage of using custom ASICs (or GPUs) for mining it relative to using a conventional CPU. Its future is even less certain that BTC, of course, as a later comer, but the technically interesting bits lie in its proof-of-work hash function: Scrypt . Scrypt is designed to be "memory
By David Andersen – August 28, 2014I woke up on May 28th, 2014, on vacation with my family in the middle of the desert, to find a copy of my private source code plastered across the bitcointalk message board. Announced as a "new optimized version" of the Monero currency miner, it was enthusiastically adopted by cryptocurrency miners across the world. And in the process of doing so, my daily profit from the Monero Mining Project dropped by over five thousand dollars per day. But let's start at the beginning, when I started getting in to a loose collaboration with three people I've never met---one whose name I'm not even confident I really know---with hundreds of thousands of dollars worth of a nebulous new cryptocurrency at stake. It started with a cryptic note from someone I'd met online, with a link to a bitcointalk.org message board discussion for a new currency called "bitmonero". His note said only: "this looks interesting." From prior collaborations with him