LESSWRONG
LW

Comment Permalink

[anonymous]14y50

Also, in my 3 years of graduate school, I have had to maintain LOTS of legacy code left over from previous grad students. This is a good example of what I mean. You are assuming that "software jobs" has a negative factor of managing legacy code, and for some reason you don't think this applies to graduate school. If you study in any applied science field at all, you will have to maintain poorly-written legacy code.

You should view applied science graduate school as a low-paying, low-benefits version of a software development job for places like Yelp or Google or Facebook. You will do Bayesian inference and machine learning on large data sets, no matter what you think you will do or what an adviser says you will do. You will spend less than 10% of your time actually investigating new, original ideas in science, and the other 90% of the time you will do the day-to-day grind of a typical software engineer.

You will have access to good facilities, lots of interesting seminars and colloquia, and the opportunity (/ huge time cost) to take interesting advanced courses. You will also have the freedom to sleep in, arrange your working hours haphazardly, and it will be socially plausible to continue "acting like a college kid" for a few more years. These are definitely good benefits.

But my experience has been that (a) future employers won't care very much about grad school unless you are an amazing programmer; (b) you only earn ~3% more than a person with only a master's degree in your same field, and since they have a ~3 year lead time on you, their lifetime earnings are higher; and (c) the charm of "intellectual student life" wears off pretty fast when you need a good dentist but don't have insurance or when you need to fly home for a family emergency but literally cannot afford the $500 plane ticket, or when you can't travel home around the holidays because there is a journal deadline.

There's just so much to consider that you won't learn if you only talk to faculty members for whom the grad student experience has worked out smoothly.

Dreaded_Anomaly14y00

(c) the charm of "intellectual student life" wears off pretty fast when you need a good dentist but don't have insurance

From personal experience, I know that the University of Maryland offers state-subsidized health insurance to graduate students. I would advise considering the availability of insurance as part of one's criteria for selecting graduate programs, rather than taking uninsured status as an inevitable result.

11

What to do after college?

9th Oct 2011

1 min read

11

My friend is looking for some advice on what he should do after graduating from Harvey Mudd College. Some relevant bits of information about him are that he

is not a US citizen, so he'd only be able to stay in the US if he's working or at a grad school. He's open to suggestions for other countries.
is great at math and computer science, including doing real-world programming
wants to help the world

He's currently looking for a grad school where he could tackle interesting problems with possible high benefits in the future. I've made my own suggestions, but I'd like to get a (somewhat) independent set of opinions from the LW community.

So please suggest away!

11

New Comment

21 comments, sorted by

Click to highlight new comments since: Today at 10:44 AM

[-]lessdazed14y130

Remember: ask tenured faculty about their career path, and they will tell you, ask those who won the lottery about their career path, and they will tell you.

Ask those who failed to become tenured professors and failed to win the lottery.

[-][anonymous]14y30

I wish I could give this more than one upvote.

[-]Manfred14y110

There's always the obvious route for helping the world with programming. Start a company, make decent money and sell it for lots of money, donate heavily to charity.

[-]multifoliaterose14y50

You might point him to the High Impact Careers Network. There's not much on the website right now but the principles have doing in-depth investigation of the prospects for doing good in various careers and might well be inclined to share draft materials with your friend.

I'm the subject of the post -- startups probably won't be able to hire me because of the legal hoops they'd have to jump through to hire a foreigner... As some of you mentioned, I'm considering writing software but I don't really enjoy maintaining legacy code.

Thus, grad school seems like a good way to develop new things (and stay here legally) What do you think are cool fields?

[-]jpulgarin14y60

There's not that many legal hoops at all. WIth a bachelor's degree it's pretty simple to get an H-1B visa. I'm from Canada and a ton of my friends get jobs at tech companies after graduation with no problem. Larger companies like Google, Facebook, Microsoft, Zynga are obviously more open to this type of thing, but many of my friends have gotten jobs at early-stage startups as well. You can also start your own tech startup via this "loophole".

You may also want to look at the expected value of being one of the early engineers at a tech startup. My prior is that it is significantly lower than the expected value of starting one.

[-][anonymous]14y50

Also, in my 3 years of graduate school, I have had to maintain LOTS of legacy code left over from previous grad students. This is a good example of what I mean. You are assuming that "software jobs" has a negative factor of managing legacy code, and for some reason you don't think this applies to graduate school. If you study in any applied science field at all, you will have to maintain poorly-written legacy code.

You should view applied science graduate school as a low-paying, low-benefits version of a software development job for places like Yelp or Google or Facebook. You will do Bayesian inference and machine learning on large data sets, no matter what you think you will do or what an adviser says you will do. You will spend less than 10% of your time actually investigating new, original ideas in science, and the other 90% of the time you will do the day-to-day grind of a typical software engineer.

You will have access to good facilities, lots of interesting seminars and colloquia, and the opportunity (/ huge time cost) to take interesting advanced courses. You will also have the freedom to sleep in, arrange your working hours haphazardly, and it will be socially plausible to continue "acting like a college kid" for a few more years. These are definitely good benefits.

But my experience has been that (a) future employers won't care very much about grad school unless you are an amazing programmer; (b) you only earn ~3% more than a person with only a master's degree in your same field, and since they have a ~3 year lead time on you, their lifetime earnings are higher; and (c) the charm of "intellectual student life" wears off pretty fast when you need a good dentist but don't have insurance or when you need to fly home for a family emergency but literally cannot afford the $500 plane ticket, or when you can't travel home around the holidays because there is a journal deadline.

There's just so much to consider that you won't learn if you only talk to faculty members for whom the grad student experience has worked out smoothly.

[-]Dreaded_Anomaly14y00

(c) the charm of "intellectual student life" wears off pretty fast when you need a good dentist but don't have insurance

From personal experience, I know that the University of Maryland offers state-subsidized health insurance to graduate students. I would advise considering the availability of insurance as part of one's criteria for selecting graduate programs, rather than taking uninsured status as an inevitable result.

[-][anonymous]14y00

Almost all science programs offer a very basic health insurance package. Almost none offer dental insurance. Even with the health insurance, you often have to pay a ~$500/semester fee plus completely pay for the summer expense. Using Maryland as an example, I know an applied math graduate student there and she has to pay $1000 towards her own insurance plus cover it in the summer and gets no dental benefits. I go to a very wealthy Ivy League school and I also have to pay a little less than $1000 per academic year for student health insurance. I can purchase optional dental insurance for $500, and all it gets me is a 5% to 15% discount on the out-of-book price for dental service, plus two annual cleanings. Further, no one in the area takes it. My university's own dental services do not accept their own university grad student dental policy.

[-]Dreaded_Anomaly14y00

Having to cover the insurance during the summer suggests to me that she does not have an assistantship, or at least not one which will continue over the summer. That is something else which must be evaluated on a case-by-case basis, and will not necessarily be a factor for every student.

$1000/year is close to an upper bound for health coverage, given the range of subsidized rates provided by Maryland. Dental insurance is available for an additional amount of around $100/year. They offer both a DHMO and a DPPO, the latter of which can be used at any provider.

[-][anonymous]14y00

She is a 4th year PhD student in applied mathematics working with an established adviser who has plenty of grant money. She's just doing the standard thing... I see the same situation all over my own university's science departments.

PPO can be used at any provider and typically only get you a slight discount for any service beyond a basic cleaning. As a grad student, I've had to get a crown and 3 fillings (arising as complications from wisdom teeth, which I had removed while a full-time engineer prior to grad school). I realize my expense is at the far end of the distribution, but that's why we have insurance. Grad students (like me) aren't in a position to handle these expenses and our insurance is not remotely adequate. I had to go into significant debt to finance my own dental work.

[-]Dreaded_Anomaly14y00

It doesn't make sense to me that she would have to cover her own insurance over the summer if she continues funded research as a graduate assistant during that time. University employees get the subsidized rates.

UMD's DPPO covers 70% of costs for fillings and 50% for crowns, after a $50 annual deductible, and up to a $1500 annual maximum. Costs could still add up, but that seems like more than just a slight discount. Obviously what insurance covers will vary from school to school, but that information is generally available online.

I'm not saying you didn't get screwed. Wisdom teeth complications suck; I've dealt with some myself. I'm just saying that's not a given for all graduate programs, and it's not necessary to assume that it is.

[-][anonymous]14y10

I have encountered several professors who explicitly looked for foreign students so that they could use the student visa as leverage. These were not bad researchers either; one is a very famous mathematician from Brown University. Several of my foreign friends described feeling very scared and depressed because if they did not publish good enough results, they faced being deported (I'm not exaggerating; maybe they perceived the threat to be more plausible than it really was, but even so, do you really want to feel like you could be kicked out of the country?)

In a few cases I have specific knowledge that these students fudged research data, left some bugs in their code, etc., just to rush their results to print and submit them to journals.

I'm not saying that it's likely this would happen to you. Just don't go to grad school because of visa issues. When I was a high school student, I visited the US Naval Academy for a three-week recruiting program. I thought of it as a great academic institution, lots of access to technology, and it was free. They did a good job of berating that point of view, and making it clear that you should only try to become an officer in the Navy if you want to be an officer in the Navy. It is, in fact, not a free education at all, but rather costly both in terms of your cultural experience and time you will serve in the Navy.

It's the same with grad school. Unless you are fully aware of all of the depressing, negative outcomes that most graduate students experience, and you still cannot possibly choose any other career path, then I strongly advise you to consider other options. Work for a financial firm with offices in Hong Kong or Singapore (like GETCO for example) and earn your chance to get transferred back to New York if life in the United States is very important for you. As others have mentioned, getting an H1B visa is not that hard if an employer wants to hire you. Most software companies will do this for a talented employee. You could also consider working in Europe where you have many of the same freedoms and luxuries of life in the US.

The overall point is you need to pretty much independently decide if grad school is right for you first and only then worry about the visa situation. Unless you have some very unusually urgent reasons to stay in the US at almost any cost, then do not just do any old job just to stay here. Find a job you actually want to do.

[-][anonymous]14y00

This may seem like an odd question, but have you been involved in any high-level math competitions?

[-]atucker14y40

I always find it weird when people don't comment on advice threads. The potential utility of significantly altering someone else's life seems pretty big compared to lots of other things.

Unless of course you don't think that there's much chance of the advice being used, or of the advice being substantially better than the next best option.

[-]Dorikka14y80

If I don't feel like I am succeeding in the area in which someone is requesting advice, I'm not going to give advice, as it may be destructive.

[-]atucker14y50

You could give reasons why you don't think that you're succeeding, at the very least.

[-][anonymous]14y30

From the information provided, I'd say: skip grad school, get to Silicon Valley, join a startup with high monetary expected value, donate to causes with high expected utility, and iterate until we are all immortal.

[-][anonymous]14y20

I attended Rose-Hulman (a school very similar to Harvey Mudd) and majored in mathematics (class of 2008). I worked for two years as a research analyst for MIT Lincoln Laboratory and then earned my master's degree in applied mathematics. I am currently in my 2nd year of a PhD program in applied mathematics, focusing on machine learning in computer vision.

I have written posts about jobs/school/future here, here, and here. I also recommend reading this and this regarding grad school.

The criteria for your friend are not very specific and I suspect that signficantly many more specific factors are needed to decide a good route forward. However, in a very broad sense, these are my observations:

(1) Do not go to graduate school, except possibly just to earn a very technically relevant master's degree.

(2) Learn ways to "make a difference" or "derive utilons from activities" through your non-professional work life. Technological work is almost exclusively the action of creating value for a profit-driven entity. Work that is not highly focused on this task will probably not be intellectually stimulating enough to leave you feeling content with what you're doing.

(3) Along the lines of (2), do not discount what money can get you. Money is not everything and there's plenty of evidence that it only weakly correlates with happiness. However, if you want to achieve certain goals, you'll need to finance them. For example, if you work at a hedge fund and are earning over 6 figures for decades, you'll likely be in a position to offer prizes or scholarships in a form of charity angel behavior. This is not a suggestion that you should chase money, but give fair consideration to the idea that you can accomplish a lot more goals later on if you spend a few years of your young life working a bit more doggishly towards profit than might seem "benevolent" by most people.

Tenured faculty positions will not exist by the time your friend is in a position to try for one. More over, this friend will have to spend 5-6 years doing unpleasant computer programming work as a graduate assistant, followed by 2-4 years as a low-paid post-doc with sub-par insurance, followed by between 4 and 10 years of time as an assistant professor, and only then will they really be able to "calm down" and have a normal life if they are awarded tenure.

If you can prevent yourself from desiring the PhD or adopting the false belief that PhD == chance to do intrepid, avant garde intellectual research, then you'll be far ahead of the field. Make your peace with doing software development / data analysis for a bank / hedge fund / consulting company / software company ... live well below your means ... and explore lots of volunteer opportunities, open source software projects, etc., to form your worldview about which domain you will try to influence later on.

I'm sure that TONS of people will disagree with me, which is OK. You definitely should investigate the alternatives to everything I am saying. This is just my life experience and what I can ascertain from studying the same exact problem after already sinking ~3 years into grad school.

If your friend does decide to pursue mathematics grad school even after reflecting on all this and reading my other posts, feel free to private message me and I can give more targeted advice about how to look for an adviser and how to choose a good fitting university, field of study, etc.

Thank you for your advice. I wasn't really sure what a grad student's life is like and given the high opportunity cost and limited benefits it seems like a bad idea all around.

It also seems like an excellent opportunity to do a bit of traveling; it can give a bit of a different perspective, and a good time to think about future plans. Many people seem to have difficulties to take extended time for traveling during their post-uni/working career (for various reasons), so this may be a good time.

More from mtaran

Curated and popular this week