(no guarantees that i'll be able to answer via text, though; maybe i'll make a video later. been trying to minimize my computering time off-hours due to increasing wrist pains ... PSA: take care of your wrists, everyone!)
Thanks so much for writing this memoir! It is absolutely brilliant. Considering you had a somewhat unconventional PhD with your independent projects, what was the main role of your advisor? How do you make the most of such a situation to extract knowledge out of professors who may not have an incentive to be directly involved in your project to the usual degree?
if you can't get on someone's critical path, then you have to make it very easy for them to help you with very little time commitment. e.g.,: http://pgbovine.net/how-to-ask-for-help.htm
What made you go back into academia? Based on this snapshot of your life, it felt like academia at its best gave you an outlet to explore the things you were really interested in, but at its worst, had a ton of obvious drawbacks. Even in this book, it seems a lot of the best experiences stemmed from or started outside of the program (MSR) and the Ph.D, while providing support and enabling these things to happen in the first place, was no longer an active contributor in.
I know you had mentioned you could write a whole book on this, so I'm sure there's a lot to the story.
you've inadvertently answered the question for me, to a first approximation :) i think that academia is a great launching point for a wide array of scholarly activities that the free market (i.e., industry) doesn't directly pay for: research, public policy, outreach, teaching, mentoring, industry collaborations, etc. i can work with whatever companies i want (even ones that are actively competing with each other at the moment!) and be seen as a "neutral" party; i can share knowledge via teaching and research again with a "neutral" voice without being seen as a spokesperson for a particular company or other special interest group. you're right, though -- there's a whole lot to the story. maybe someday i'll write something up!
It's a good read but it is also worth knowing that Dawson's group was going through changes at the time and PhD experiences are often very unique depending on the circumstances of your advisor, research, and field trends.
totally agreed! things can change even from one year to another with the same advisor. for instance, i'm not the same advisor to my students that i was last year, or the year before that, or the year before that. (i've only been at this for 4 years, and each year is incredibly different from the prior one.) circumstances change, resources change, and constraints change.
I was grad council president at that Harvard physics department a couple of years ago, and I helped organize a similar survey.
My first impression here is that things aren't too bad. Some people are commenting, for example, about projectors not having enough cables --- obviously a nuisance, but if those are the issues the department is having, great. (Obvious, easy solution and a relatively small problem to begin with)
Second, I don't really think this is a particularly comprehensive survey. I'm surprised there weren't more pointed questions about job intentions (academia vs industry vs not sure), feeling respected by those around you, the prevalence and (separately) severity of the racism, sexism, and general harassment, etc. Not a single question about teaching, nor a question about actions they'd like to see taken. Relatively few culture questions outside of "do you feel like a community" (something like that).
At Harvard physics, there were more pressing concerns regarding sexism, racism, the quality of the required courses, and professional development. The department did reciprocate and make great changes to the required courses, but deeper biases make the other categories much more difficult to tackle.
I have been helping a lot of students deal with considering industry, as in a physics department it's looked down upon to just be considering industry, let alone pursue it. Or, even if not explicitly looked down upon, it's definitely the culture and general feel, and many students and postdocs prefer to discuss it in private. (I did an internship at Waymo last summer, so people are especially keen on asking me about transitioning to industry)
The survey results are unfortunately not public, so I can't share them. Kudos to the group of students or administrators that put together this data! I'd love to hear what they learned from it and changed as a result of it.
I did a Master in CS, worked as a developer for 5 years, then went back to do a Ph.D. I stayed with it for a year before deciding it was not for me, and went back to SW development.
My key reasons for not continuing with the Ph.D.:
- Many problems you study are chosen because you will be able to publish something, not necessarily because they need to be studied.
- You don’t need to be a (Ph.D.)student to learn - you can work and still learn
- A Ph.D. in itself doesn’t make you smart
- Narrow problems vs broad problems - I prefer to work on something where all parts need to be good enough, vs on finding the best possible solution to a very narrow problem.
- Having worked before starting the Ph.D., I could compare working in industry vs studying for a Ph.D., and I realized it was very stimulating in industry.
As someone with a CS master's I almost wish there was an alternate PhD path where you didn't have to do research, like a Masters++ program. There were many classes I never got to take for my degree, like in category theory or AI or distributed databases.
Some of the PhDs in this thread have said they tried $BigTechCo and found it boring, but to me that just means you were working at the wrong company or on the wrong project.
I feel like my master's equipped me to be able to read and study theoretical computer science at a high level, that the PhD program would be a strain on me and my family without much reward.
Although I do daydream about being independently wealthy and getting a master's in mathematics just for fun
This doesn't answer the question I fundamentally want to know: do they think it was worth it?
In my CS master's programs profs often warned against the PhD, said companies wouldn't hire you, that they'd view you as "too smart" and potentially be bored by everyday work.
Not true for certain disciplines (AI, e.g.) but I don't think every CS PhD is walking out the door with 300K/year offers hitting them in the face.
Granted that no one does a PhD because they want to starve, but I think looking at it strictly from an instrumentalist point of view (i.e. will this bump my salary enough to offset being paid peanuts for 5-6 years?) is not very useful. Almost anyone good enough to get into a PhD at a top tier school is also good enough to make buckets of money without one. If you want to make money, go make money.
A PhD might lead to money, but it's hardly the easiest or most rewarding way to do so. You generally finish classes after 1-2 years -- after which point you have a broad overview of problems in your field and a few more years of practice and professionalization -- and the rest of the time is just researching and pushing the frontier on a narrow, deep problem of interest. Some of the frontier pushing is useful in industry, some of it is purely academic. If you finish the PhD and aim for a tenure track position or a few postdocs, you can continue that type of work. If you exit to applied industry, the type of problems you'll get to work on are very different.
I think you should consider a PhD if you enjoy tinkering with problems and pushing your intellectual boundaries, don't particularly care about money, and can see yourself spend several years grinding at something to produce what amounts to a book on the subject (note: a real book, not a "57 Tips To Make Sites in React" ebook to plug on medium blog posts)
In terms of whether companies are bothered by people who have too much education? Sure, some might. Certainly a PhD is not prepping you for the kind of work conditions that exist in tech startups. There's no reason to believe a PhD will be notably better at a problem than someone with an MS in the same field, and if they have 5 years less real-world experience, then there's that consideration too.
You do a PhD if you are truly passionate about the work. You do not do a PhD for money, for admiration, for anything else. You do it because you have to know something about something, and you have to know so much about it you discover something no one else knows about it, no matter how small.
There is absolutely no other reason to get a PhD because while there are lots of perks to a PhD - traveling, meeting people who share similar interests, interesting career opportunities, autonomy, collaboration - etc, you can find all of those perks at a different job that isn't so intellectually demanding. The one thing no other job can offer you is proof that you discovered knowledge that is actually new (or whatever way you want to phrase it, there exists no other institution outside of academia that is rigorously structured to ensure you don't get a PhD without doing this at least once).
You can say this is an argument of authority controlling a body of knowledge, I am sure that will be the typical start-up perspective. Unfortunately, a start-up's bottom line is to produce a product. That means that if it comes down to the mark of determining the integrity of some knowledge, or keeping the business running, the start-up is not going to destroy itself for the sanctity of knowledge. Plenty of PhD students, however, do!
You literally will have no life besides your work if you pursue a PhD. There is zero reason to do one unless you have this drive.
I have a CS PhD from Stanford (graduated 2014) and it is definitely worth it. The thing it actually buys you is freedom to choose your own research, which you can't do with a bachelors or a masters. I now work at an industrial research lab where I get to set my own research agenda within some reasonable bounds of interest of my employer. I pick the projects I work on and am compensated based onto how valuable they are to the company. Importantly failed projects can be just as valuable to the company as successful ones as they can show the product groups what NOT to do. The PhD is effectively a license showing that you've been trained to conduct research. Most companies won't give you the freedom to pick your own research topics without it. For someone that cares a lot about what I'm working on, that freedom is more valuable than money.
This ultimately depends on your definition of 'worth it'. The primary reason to do a PhD is to do research. If you know you want to do research (here defined as high-impact work publishable in top journals) as a career, then a PhD is probably a good move. I've worked across a few industry research labs, and 95% of the folks have PhDs. There are some exceptions (one of the best researchers I know, with some of the highest cited papers in the field, has a BA in music from a no-name Canadian university), but by and large getting a PhD is the most direct path to a research career.
If your definition of 'worth it' is purely financial, then you are probably better off hitting the job market with a BS or a Masters.
For foreign students, the PhD program is a great way to enter the U.S. job market. In fact, I'd go as far as to say that, while American undergrads can get lucrative job offers right out of college, foreign students are left with doing a Masters/PhD first, hence partly explaining a skew towards foreign students in many Computer Science departments around the country. For those students, I'd say the PhD program is "worth it".
Disclosure: foreign PhD candidate at an American university here.
I don't think Google has ever turned away an applicant for being "too smart".
That being said, five years in a PhD is probably not as good for your career as five years' experience at $BigNameTech. PhD is probably more fun, though.
I have many friends in PhD programs, including CalTech, Stanford, and Berkeley, pursuing CS, Genetics, Biomededical Engineering, and EE.
To characterize their program is “fun” is a gross mischaracterization, and demonstrates a serious misunderstanding of the rigors involved in any technical/science/engineering PhD pursuit.
PhD programs are hard. They’re often lonely. They’re frustrating. They’re tedious. And there’s a tremendous amount of pressure. And it’s academia: highly bureaucratic and political and often unfair.
Unless you possess a serious passion for the subject, you’re unlikely to survive.
On the other hand, my friends getting the PhDs in subjects such as education (Stanford) and conservation biology (Mississippi State) have described it as the highlight of their life. After their first two years, they’re basically getting paid to do whatever they want. They audit classes. Apply to grants. Travel. And they write.
The irony of this reply is that I did do a PhD, in pure math at a top 10 school, and while I might describe it as hard, lonely, frustrating, tedious, and high pressure, I would also describe it as having been fun. Certainly more fun than my current job at $BigNameTech.
Thanks for the lecture about my Serious Misunderstanding, though.
Do you regret doing your PhD in math? Reason I ask is that I thought about going to grad school for math but decided against it. It just seemed like it would be postponing the inevitable of finding a job that had little to do with my "passion". I definitely see the appeal of doing a PhD in a field with better industry employment prospects (e.g. AI in CS if you're into that), but less so for fields like pure math that aren't as employable outside of academia. I'm curious if you felt the experience was worth it, though
I did a PhD in math and if I were to do things again, I'd do applied math because I think I enjoy applied math at least as much as pure math (this is not true of everyone) while also having it have more obvious job prospects. I did pure math in part because I was told it was easier to go pure->applied than applied->pure. I still think that is true and was a not an unreasonable way to choose what to study. I fundamentally had a very good time in my PhD (I would absolutely describe it as fun, though that would only be part of the picture) and learned a ton and now am transitioning to applied math in exactly the manner that I had been told would be possible (though I feel like I did this more by luck than by it being systematically possible).
I wouldn't say I regret it, but that almost doesn't feel like an applicable word; it felt inevitable to me, so it's hard to imagine any other path. Like, at the end of undergrad, all I wanted to do was think about math all day, every day, and that's what I got to do for the next five years. There was no convincing 22-year-old me to do anything else. Only near the end did the novelty really start to wear off.
N.b. I'm in a very lucky position career-wise now. If that were not the case, maybe I would feel more regret.
Yeah, fair enough. It's more like it could be more fun, if you're obsessed with the field, and your advisor isn't terrible, and your program doesn't encourage cutthroat competition (by e.g. admitting more students than the faculty can realistically support), and you're willing and able to live a Spartan lifestyle (in particular, no financial dependents), and ...
I'm lucky in a lot of ways and I sometimes forget that.
I think your previous answer using the word "fun" was spot-on. I have a non-CS PhD that's not really related to what I do today. I would describe what I did previously as "fun"; despite the intensity and worse WLB than life in a big-tech company, you're ostensibly choosing projects that you and your adviser consider fun. In bigtech, you have a lot less freedom to choose what to do. Seriously, if you don't consider your PhD fun, you should just drop out now; if you plan on just going to industry later to make money (esp as a SWE), there's no reason to get a PhD.
It’s interesting, of course, that a similar set of caveats apply to working for early-stage startups - not that this forum would attack anyone for calling working for one fun.
I got a CS PhD at Caltech. It was fun. Experiences vary a lot with who your advisor is and what topic you're researching. The work could be frustrating and tedious, but I never felt overly pressured and the people running the bureaucracy always just made problems disappear for me.
On the other hand, my friends getting the PhDs in subjects such as education (Stanford) and conservation biology (Mississippi State) have described it as the highlight of their life. After their first two years, they’re basically getting paid to do whatever they want. They audit classes. Apply to grants. Travel. And they write.
I wonder if they have family money.
This is actually not a snide comment. The people I knew who did a PhD and liked it had enough money, typically from families and sometimes from a spouse, to not worry about money, or five+ years at substandard wages. I worked part-time as a grant-writing consultant, so I was fine, but I saw financial privations all around me. To the point I wrote this: https://jakeseliger.com/2012/05/22/what-you-should-know-befo...
Granted, science PhDs are often better funded and more lucrative post-degree, but even among people in the sciences financial factors often loomed.
Not a snide comment. I had two friends who entered into PhD programs, one for Math, the other for engineering.
Both dropped out for financial reasons.
One was married and was getting a lot of pressure from his new wife about starting a family, settling down, getting a real job that pays better than what the college was paying him.
The other had been on his own since he was 16. Bad family situation, he was raised by his grandparents. One of the sharpest guys I knew. Just burned out from trying to work three jobs and keeping up with the program. It was just too much and he needed to get a job that paid him better so he didn't have to work 70 hours a week to get by. The funny part was five years later, he landed at Google, was making a lot of money, and was planning on going back with help from Google to finish his degree.
> Unless you possess a serious passion for the subject, you’re unlikely to survive.
I mean, should you really get a PhD without a serious passion for the subject? I won't disagree one bit that it's not for everybody.
Speaking as someone wrapping up an engineering PhD from a good school though, I enjoyed it quite a lot and I think most of the people I work with enjoyed doing their PhD. It was not easy, but it was interesting and rewarding. The environment varies a lot depending on the school and the adviser. Some programs are indeed quite toxic, but there are also plenty that are challenging, but fulfilling and nurturing.
> I mean, should you really get a PhD without a serious passion for the subject?
I'm not sure why passion is particularly important. I would rate work ethic, diligence, attention to detail, and some character (i.e., not lying in studies) as more important than passion.
Generally, making people jump through hoops to get something done doesn't actually get you the best people. It gets you the people who are good at gaming such systems, and people who don't know what they are getting themselves into.
A PhD is a pretty deep dive into a subject, if you don't have passion you're gonna burn out. Of course other things are important, but if you don't have passion for the subject you're gonna be miserable.
And I don't really see it as jumping through hoops. A good PhD program has a few checkpoints to pass, but everything else is up to you. I strongly disagree that you don't get good people, at least not the ones who finish. People who don't know what they are getting in to or are only good at "gaming the system" tend to wash out pretty quickly, usually because they are miserable and burn out. It varies from industry to industry, though... maybe other engineering fields are worse than the ones I'm familiar with.
A lot of people make the mistake that you have to have a PhD to prove you're "smart", which is wrong. But that doesn't mean that only people with nothing better to do get one.
>PhD programs are hard. They’re often lonely. They’re frustrating. They’re tedious. And there’s a tremendous amount of pressure. And it’s academia: highly bureaucratic and political and often unfair.
>Unless you possess a serious passion for the subject, you’re unlikely to survive.
This is mostly a recent phenomenon in engineering. I went to a top school and left almost a decade ago. While advisors like what you described existed, roughly half of them were not like that. It really was fun, and relatively relaxed.
These notes[1]--written by Prof. Comer of Purdue--and over a year of introspection ultimately led me to decide against the PhD path (EE, not CS). To be sure, the words fun, happy, and rich were used precisely zero times. It's a humbling quick read for anyone else out there considering.
As a PhD student, I'll lightly disagree and say that I'm happy in my PhD. I like doing research every day and having freedom to work when and where I want, on what I want. It's much better than the year I worked as a software developer working on things I found boring and repetitive. I would definitely say I've had fun. Of course, it's not always fun, and I know I've been lucky as far as advisor and project randomness. But the harshness of the link you posted seems unnecessarily severe.
The notes are very apt, but it's worth noting that a PhD is so fun that the people who finish do so in spite of all the misery described in that essay. It's the best and worst decision of your life.
Being a PhD student in something like Stanford, MIT, Berkeley can be fun in some way, but in some other ways it's mind blowingly stressful. Working in $BigTech can be boring, but it's usually almost 0 stress.
Stanford education comes with access to Stanford faculty and students. These are some of the most accomplished and talented people in the whole of Computer Science. It certainly is in a different class of benefit than job offers and you cannot always put a K/year value on it.
I haven't pursued a masters degree before, but if I did it wouldn't be for the money. I could never stay involved enough to be successful.
I am still intrigued by the idea of a CS PHD but I stopped after my Master’s. I cannot emphasize enough the importance of finding an advisor you mesh well with.
CS PhD student (not at Stanford) here. Based on the survey timing, 2015 respondents were in their 3rd year when answering this survey. I have personally heard and read that 3rd year is a hard year for many PhD students (and the experiences of my peers and I largely bear this out), so the dissatisfaction here tracks with that idea.
There are many possible explanations:
1. Most people have their master's degree by third year. There's a sense of "if you're going to drop out, now is the time to do it." If you're miserable in your program but hate quitting stuff, third year might be the year that finally breaks you.
2. Coursework is largely over by 3rd year, and a student should be doing research close to full time. This can be a hard transition. Granted, most Stanford students probably have substantial research experience coming in, but even that is not the same as doing (often very unstructured) research all day every day.
3. The "honeymoon" is over. You're no longer a young student, and pressure is growing to publish, know your area, network, and so on. At the same time you're still quite junior, so you know you're probably not very good at any of these things yet. This can be a frustrating combination. Kind of like adolescence. Also, if you've been unlucky with conference reviewing, you may have a stack of 2-4 papers that have been rejected at least once or twice, and you despair of ever doing anything externally recognized as useful. In the other direction, some of your peers now have half a dozen accepted papers at good conferences, and you feel inferior (never mind that these are small sample sizes, and peer review is noisy).
So some self-selection occurs in the third year, and the group that sticks around to year four is usually smaller and happier (and, of course, some happier fourth years used to be miserable third years).
Thanks! This is definitely making sense to me. I'll bookmark this and hope they do another survey next year (and share). Looking forward to the comparison of year 2016 next year.
I recently emailed somebody a few quotes from Freeman Dyson on the PhD system, which I include below. I can track down the sources when I'm back in my office, although anybody can find them easily enough. Dyson is not obscure.
“I’m very proud of not having a Ph.D. I think the Ph.D. system is an abomination. It was invented as a system for educating German professors in the 19th century, and it works well under those conditions. It’s good for a very small number of people who are going to spend their lives being professors. But it has become now a kind of union card that you have to have in order to have a job, whether it’s being a professor or other things, and it’s quite inappropriate for that. It forces people to waste years and years of their lives sort of pretending to do research for which they’re not at all well-suited. In the end, they have this piece of paper which says they’re qualified, but it really doesn’t mean anything. The Ph.D. takes far too long and discourages women from becoming scientists, which I consider a great tragedy. So I have opposed it all my life without any success at all.”
From a different interview:
“Well, I think it actually is very destructive. I'm now retired, but when I was a professor here [Institute for Advanced Study in Princeton], my real job was to be a psychiatric nurse. There were all these young people who came to the institute, and my job was to be there so they could cry on my shoulder and tell me what a hard time they were having. And it was a very tough situation for these young people. They come here. They have one or two years and they're supposed to do something brilliant. They're under terrible pressure — not from us, but from them.
So, actually, I've had three of them who I would say were just casualties who I'm responsible for. One of them killed himself, and two of them ended up in mental institutions. And I should've been able to take care of them, but I didn't. I blame the Ph.D. system for these tragedies. And it really does destroy people. If they weren't under that kind of pressure, they could all have been happy people doing useful stuff. Anyhow, so that's my diatribe. But I really have seen that happen.
And also, of course, it wastes a tremendous amount of time — especially for women, it's particularly badly timed.
If they're doing a Ph.D., they have a conflict between raising a family or finishing the degree, which is just at the worst time — between the ages of 25 to 30 or whatever it is. It ruins the five years of their lives.
And I see the difference in the business world. My daughter happens to be a businesswoman, so I meet a lot of her young friends.
The life there is so much easier for women. They start a company when they're 20; they go bust when they're 22. [Laughs] Meanwhile, they have a kid, and nobody condemns them for going bust. If you're in the business world, that's what's expected: You should go bust and then start again on something else. So it's a much more relaxed kind of a culture. It's also competitive, but not in such a vicious way. I think the academic world is actually much more destructive of young people.
[The Ph.D. system] was designed for a job in academics. And it works really well if you really want to be an academic, and the system actually works quite well. So for people who have the gift and like to go spend their lives as scholars, it's fine. But the trouble is that it's become a kind of a meal ticket — you can't get a job if you don't have a Ph.D. So all sorts of people go into it who are quite unsuited to it. [...]
Anyway, so, I'm happy that I've raised six kids, and not one of them is a Ph.D.”
CS and engineering PhDs seem to fall into a kinda weird category of "industrial" PhDs where it seems more people go in from the get-go with no interest in staying in academia. I think some CS/engr programs actually have built-in expectations of doing internships with industry, which seems really backwards to me.
> I think some CS/engr programs actually have built-in expectations of doing internships with industry, which seems really backwards to me
Why do you think this is backwards? Industry often has more resources for research than academia does, so doing an industrial internship is usually a way to supercharge your research and get better data and try more things.
Coming from a non-CS background, the fact that the summer is often not a great time to get work done with CS graduate students because they're away at internships is taking a lot of getting used to.
Depends. If you are American a bachelors is enough but a masters can help. If you are a foreigner who wants to work in America then a Masters helps (1 year of school in the US) and gives you two slots at the lottery. Machine learning and data science jobs that are relevant in select companies usually require a PhD in math/stats/cs/engineering especially as you go up the food chain and corporate hierarchy. CS is a little more forgiving if your are adept at developing complex software systems. This is what I've observed at a few public companies that are relevant in machine learning / data science.
Grad school can also help accelerate your career especially if you can program.
I much prefer divergent stacked bar charts for likert/1-5 ratings, rather than these side-by-side bar charts. It's really hard to compare the distributions. Even a box plot would be better, I think.
On a different note, it would be interesting to know how many respondents opted for the eGift card for completing the survey. And if the responses deviated a lot when the data is pivoted with that variable.
Curious. Stanford sounds like it needs to bring in adjunct ML faculty from industry to serve as academic advisors. Also their weed out "breadth" classes sound dumb, most would hate the program.
http://www.pgbovine.net/PhD-memoir/pguo-PhD-grind.pdf
https://news.ycombinator.com/item?id=4179982