The N=1 Guide to Grad School: Advice for Aspiring PhDs | Hacker News Reader

The N=1 Guide to Grad School: Advice for Aspiring PhDs

marcua.net

124 pointseob39 comments

It's what he's doing now that doesn't mesh with me... doesn't working at Locu feel like a huge step down in terms of how interesting the problems are?

I've always been worried about being spoiled with ideal situations and infinite time to read and catch up in CS.

"infinite time to read"

Depending on what you do (and where you do it), the private sector can actually be quite liberating. I find academia stifling.

Could you elaborate? I mean I agree about academia being stifling, but I have trouble imagining it easy to find a research/reading sort of job in CS today.

I've run into people with great private-sector positions, but they seem to almost all be ex-academics who first established themselves in academia, and then jumped to something high-level and researchy, like a position at Microsoft Research, or a senior position at Google with considerable freedom [1].

For a junior person, academia does have a lot of downsides, not least because of the massive move towards rationalization, metrics-counting, and a focus on how much grant money you bring in. But the junior people I know in the private sector have even less freedom regarding what they spend their day doing, and much less freedom when it comes to formal working conditions, such as choosing to work from home or a coffee shop a few days a week.

Is there a way to find positions with research freedom in the private sector without first becoming established in academia?

[1] Matt Welsh is a good example of someone who did the senior-academic-to-industry transition. He went into academia, got tenure, got promoted up the ranks through to Full Professor at Harvard, and then left for a senior position at Google with considerable freedom: http://matt-welsh.blogspot.com/

"Is there a way to find positions with research freedom in the private sector without first becoming established in academia?"

Define "established"?

Most relevant to the hypotheticals here, one question is whether you can get a position with research freedom in the private sector without going into academia at all. If you choose industry rather than academia, is it possible? Or do you need to first get that "professor" line on your resume to be able to land those kinds of industry positions?

I work for IBM Research. While I do not have complete freedom, I feel that my research and development mandate is quite broad within my project. I came straight from grad school. Most people that are hired into IBM Research were not professors, but are coming from grad school or post-docs.

Didn't they just shutdown Hawthorne, move everyone in software to Yorktown, and lose a lot of people to Google New York in the process? What's left?

We moved, yes, but I'm not aware of anyone who left because of the move. I'm actually happier up in Yorktown - it's a better building than Hawthorne.

Anyone with a Masters degree (working as a researcher)?

Kinda. There are so e people who have Masters degrees who work in Research, but they don't typically have what I think of as a "research mandate". They have more development responsibilities. But, there are some people with PhDs who are the same.

Most people, though, have PhDs.

I strongly believe the "professor" title is not a point of strength on one's resume vs solid industry experience. In general, if one is taking the time and effort required to become a tenured professor, he/she should stick with it. Sure, there are anecdotal cases (Harvard profs going to Goog...), but those are anecdotes and not the mean.

I've wrestled with this in the past. As a CS PhD student, I will be working full time as a web developer when I graduate.

For me it came down to either (sweeping generalisations):

1) Continuing research on interesting problems, or perhaps working for interesting companies like the Bearau of Meteorology or the CSIRO (the Australian governments industry research organisation). The problems would be intriguing, and you would feel very satisfied with what you have done most days at work. However when you get home, I imagine you would be pretty mentally exhausted, and may not find much time or energy to work on your own projects.

2) Working for a web dev company, where the problems are (not always, but often) relatively straight forward, and don't take too much mental energy to solve. Sure, building business apps is not the most satisfying job for everybody, but the people are great, the company is great, you get looked after well, etc. The best bit is that when you get home, there is probably more mental energy to work on your own side projects that interest you.

I like contributing to Open Source projects, and I also find that after a difficult day at work or uni, I don't have the energy to punch out an hour of coding when I get home. So for me, I'm currently leaning towards option 2.

There are more than those 2 options.

Such as?

3) Use your R&D skills to create a useful technology or service in an area that interests you and either license it or found your own company around it. Quite a common route for academic types.

I'm familiar with a few people who have done this, and they really enjoy what they do. I wish anyone doing this all the best.

However it suffers the same problem as 1). That is, you may not have the energy for other projects. I understand the sentiment of your suggestion, which is: "My side project is also my work", but then you are committed to that project for X amount of time. Whereas with a typical side project that I might play with on the weekend, or with open source contributions, I am free to dabble with whatever interests me on that particular day. It is this freedom and flexibility that I am seeking in my life (partially because I have the attention span of a newt :)

Oh of course. I didn't mention:

3) One day I'd like to own a caravan park.

And indeed, they were complete and utter generalisations. I'm sure there are people in interesting jobs, who come home and are able to work on interesting side projects. Either because their jobs are interesting but not brain-draining, or because they are particularly driven, or any number of reasons.

I have just found that at this point in my life, those are the two prospects that I am weighing up.

I work at Locu over here - I was going to tell Adam his writeup made the front page so he could respond (I don't think he was expecting this), but I think he's asleep on the East Coast, so I'll post this before this falls off the front page. Hacker News is not the target audience of our merchant-facing product, but it is certainly the target audience of our technology, such as our local business API at dev.locu.com :)

Adam is our Director of Data over at Locu. Our operation is powered by a very unique data structuring system, crucial to which is a heavy crowd computing component. Adam's research was in crowd-powered databases at MIT, and Locu seems to follow from his research nicely. (I briefly worked in his advisor's lab when doing my masters at MIT as well.) As best as I can tell, he is not bored yet. If he is, he is hiding it well.

Here's a blog article he's written for us about our crowd computing process: http://blog.locu.com/post/25389032496/the-locu-workflow-a-cr...

And here's a more general article about our technology: http://blog.locu.com/post/36602971835/7-reasons-to-nominate-...

I'll see if I can't get him on this thread if it's still alive tomorrow morning.

My N=1 advice for aspiring PhDs is: you probably shouldn't go. Outside of CS, the prospects are shitty for all but the rockstars; might as well save yourself 5-10 years of slave wages by not getting the degree.

Not just CS, there are a handful of disciplines doing well. For instance, there is a huge shortage of Accounting academics these days. They're basically writing their own ticket.

"Hi, I am a PdD in Accounting!" Hmmm ... Sounds a little unexpected.

You're just proving the point that there's a huge shortage.

Certainly Enron et al were in big shortage of accounting PHDs. Or maybe was an excess?

Can you talk more about this? I hear more and more PhDs aren't worth it, which is really sad. What's a guy who wants to continue studying to do?

A big step for me was realizing that I don't need a university to keep learning and discovering. There's never been a better time to be an independent learner. If there's something you want to learn, then learn it. If you've got something to say, then say it online. And if you're interesting, people will listen and engage with you.

A PhD is training to be a professor, so don't do one unless that is your goal and you know how long the odds are and what the job of a professor really is. Plenty of people realize how hard it is to get a tenured faculty job, but I don't hear many people talking about how crappy the job of being a professor is for someone who primarily wants to do research. I routinely got emails as early as 6AM and as late as 1:30AM from the same faculty members while I was a PhD student; it seemed like my rockstar professors, well past middle age, were still working 80+ hour weeks to do their administrative duties, and teaching, and grant writing, and student mentorship, and still have any time left for their own research.

there's lots of online resources to keep learning (coursera, udacity, mit open courseware, etc), but the real benefit of the phd is the structured environment and the experience of a mentor to help guide you. Getting a phd just cause you want to learn more is like joining the army because you want to get into shape.

If you are really self motivated and passionate you will enjoy your phd experience, though at that point you probably don't need a phd to begin with...

Despite these cons, going through a phd is amazing (in my experience). You're surrounded by lots of smart people studying the same thing (or close to it), which is significantly different than just relying on those who reply to your forum posts. When i go into the office and talk shop with students/other professors/my advisor i get to talk details (math, different approaches, similar work, etc) as opposed to having to setup the problem over and over again, simplifying the idea and methods... Ideas move fast, and with a good advisor and research group it's easy to stop caring about who is on what publication and be excited about contributing to a solution

Edit: there are a lot of people who take the stance that doing a phd is working slave labor for a professor over the course of 4-5 years of your prime - in a sense they're right, but then again, if money is all you care about then by all means don't do a phd...you won't enjoy the experience...there are lots of other ways to make money if that's what makes you happy

> there are a lot of people who take the stance that doing a phd is working slave labor for a professor over the course of 4-5 years of your prime

There are many benefits of doing a PhD, but I couldn't get past the enormous feeling of opportunity cost. I saw enough graduating students and multi-stint postdocs failing to "make it" in academia and struggling to transition out that I realized it wasn't a road I could afford to go down. Better to start a new career at 26 by choice than at 36 out of desperation.

Also, the money issue does matter. You might think it's worth radically diminished wages it to live the "life of the mind" (I certainly did at one time), but be aware that it may have significant long-term impact on your earnings. After a couple years of seeing the lifestyle and economic milestones of friends who started careers after college I felt like I was being left behind socially and economically in enough ways that I might never be able to catch up. The contrast between successful college friends with my older academic friends "aging out" of the academic system into financial chaos and despair couldn't have been more clear.

The whole experience has made me somewhat anti-credentialist in my outlook. Since leaving with a masters and changing fields I've met a lot of genuinely smart people that I never would have run into while living in an academic silo. Now I tell everyone to learn whatever they want, whenever they want, and don't be intimidated by anyone's PhD in bullshitology. Because being an expert in some tiny ultraniche does not make them smarter or better than you.

> When i go into the office and talk shop with students/other professors/my advisor i get to talk details (math, different approaches, similar work, etc) as opposed to having to setup the problem over and over again, simplifying the idea and methods... Ideas move fast, and with a good advisor and research group it's easy to stop caring about who is on what publication and be excited about contributing to a solution

It sounds like either your dept is a lot less cutthroat than mine was, or you're early enough in the program that the novelty hasn't worn off yet. Regardless, I'm glad it's going well for you so far.

I don't agree, a PhD can be great fun and interesting. Perhaps it depends on where you do your PhD, and in what field (the UK-system, at least in many fields, affords you more freedom, as you only work on your own research... (assuming you have funding)).

And yes the hours may be long, but they are your hours, working on your self-defined project; and they are only as long as you want them to be (contrary to the private sector where 10+ hour days often are the norm, and leaving when you feel you've been productive for the day, even if it is past 6pm, leads to questions).

And don't forget the social life; shared dinners, clubs, balls, meeting interesting people on campus, not being too tired in the evening to hang out with friends. Great conversations into the night...

Also if you want to go for a walk in a college garden in the afternoon, or go to an interesting lecture at 11AM, you can. No questions asked. You can plan around your most productive hours. And reading, and developing your skills, is considered work here. While in the private sector you often need to do that after your 10-hour work-day...

So yes a PhD is not bad. Postdocs, and the tenure-race that ensues may be. But the PhD certainly is not bad...

The major problem with Academia (actually the Academia subset where I'm unfortunately stuck in) is entirely different.

Here it is very rare that people above grad students and perhaps postdocs do any research at all. They're just managers, except in rare occasions or in the very best places. Naturally, it is extremely difficult that anything good science-wise comes out of this.

The structure is frighteningly similar to a corporation, except that corporations usually end up dying if they're inefficient, whereas research institutions only need to produce papers to get funding—and that's an easy game to play.

I can hardly think of Professors in say a Math dept. managing their students. Yes, they're advisors, but they also conduct their own research.

You can often tell how much research a prof does by the size of their group. If they have 2 grad students and no post docs, they're probably active in a lot of research. If they have 45 grad students and 15 post docs, they're probably sitting on committees, writing grants proposals or are travelling to conference for 99% of their time. That's not necessarily a bad thing either, there is plenty of good work done by profs in a management role.

Not the same subject area, but this is why places such as Janellia Farms [1] restrict the size of their research groups. A PI with more than 6 people under her is considered a manager rather than a researcher.

[1] http://www.hhmi.org/janelia/

Well, Janellia is actually in my research area, so you pretty much nailed it.

Here is a guide to grad school from the advisor perspective: http://muratbuffalo.blogspot.com/2013/04/my-advice-to-my-stu...

This was interesting to read. But I would have liked to have learned a bit more about the "politics" involved in an academic career.

My experience (not in CS) is that grad students don't really have to worry about politics.

Can't speak for the entire academic career, but in my CS grad school, for students and post-docs politics did not exist. Maybe it kicks in later when going after tenure or trying to win a proposal over 500 other submissions.

A properly-run academic department insulates its students as completely as possible from politics. If, as a student, you're having to concern yourself with departmental politics, it's a sign that your advisor isn't doing their job.

That said, while your advisor's job is in part to protect you from whatever politics are going on in your department, it is important that you not be completely politically naïve when you finish.

So your advisor probably shouldn't try to keep you totally in the dark if there are political shenanigans going on in your department or university... but they should do what it takes to ensure that those shenanigans are "not your problem." In other words, said shenanigans should be educational (possibly even entertaining), but not distracting, stressful, or have any other impact on your primary purpose in life: getting your dissertation out the door.

Reply on news.ycombinator.com