Ask HN: PhD, Startup, or Career?
To add further confusion to my life, I've also been given a great idea for a startup and have the chance of working with a really cool co-founder. I really don't know which way to go with this one, any ideas?
I'll give a short description of the startup just in case you are wondering. My friend did a year long industrial work placement at a finance company while at University. He worked in the anti money laundering division. While he was there he had to use some god awful enterprise software that made use of some pretty bad pattern matching algorithms. Basically, he wants to do a startup with me where we would develop a SaaS based application to do pattern matching against bank transaction records to spot fraud, this is a big market with a few well established players. There is no doubt that we can make better software with better pattern matching algorithms, our problem is getting clients (same with everything I guess). The software is commonly refered to as AML software in case you wanted to do a small amount of background reading on it. (http://en.wikipedia.org/wiki/Anti-money_laundering_software)
67 comments
[ 1.3 ms ] story [ 136 ms ] threadafter that, figure out what you want, and follow that. you have time and little obligations (i'm assuming), so even if following what you want to do leads to a dead end, you have time and flexibility to get past that that you won't later in life.
At best, you would be wasting the extremely valuable "get out of industry free" card that an as-yet-unearned graduate degree represents. If you find yourself two years from now, trapped in an industry job which you hate, or in the aftermath of a failed startup, or -- worst of all -- laid off in a really down economy, you can go back to school then, reinvent your resume, revisit academic fields that have made progress in your absence, and meet an entirely new set of fellow students/potential co-founders.
At worst, if you go to grad school without sampling the actual world you could spend another half decade or more getting an advanced degree in a field that, once you graduate and the rubber meets the road, you might discover that you hate working in. Academic computer science and the software industry are rather different things. Being a grad student and being a professor are rather different things. Shop around a bit before committing yourself too hard to any of these paths.
Yes, there are industry research opportunities. They're a lot of work to get, because you have to get a Ph.D., and then you have to compete with all the other folks who want them. And the thing about careers is that there's no guarantee that the one which takes more work to get will be more enjoyable for you, or more renumerative for anyone, than the one which you could have gotten with just a college degree (or, as folks sometimes argue, a high school diploma -- I think that takes the argument a bit too far myself, but it worked for Jobs and Gates). The only way to know which career is right for you is to sample them. And while I'm not sure of the optimal order in which to sample them, trying the difficult, poorly paid, emotionally exhausting route first is probably not the optimal order.
The key here is that you've got to fail quickly when you're young. And no matter how smart you are, a PhD is simply not a fast form of failure -- it takes years just to get a sense of what the research life is like, let alone know if you'll be any good at it. A PhD is something to pursue after your confident that all other, faster options are deemed undesirable.
The only exception I'd make, is if you're one of those rare people who knows that you want to be a professor, with no doubts whatsoever; in that situation, you should get into the best graduate school possible, as soon as possible, and never look back.
That said, if you're asking the rest of us what you should do, then you don't really want to be a professor.
You're right that there are a few of those in every class, but the average smart person takes much longer to figure out that a PhD is a soul-sucking waste of youth and vitality.
Doing a PhD worked for me as a way to find out what I really enjoyed, by investigating all sorts of avenues (unrelated to my actual topic, which I soon hated), after aimlessly doing as I was told in undergrad and not really knowing what I wanted to do.
You don't sound like you're in the same boat at all. Maybe you have an incredible project that will fascinate you for 3-4 years, surrounded by equally passionate people; I can't tell. Equally likely you will grind it out for the last year or two, regretting not doing that start-up or something else.
To speak purely from experience, I would have to say that I had similar choices coming out of my undergrad. I went directly into a PhD program. The first two years I hated my life, and I would have agreed with you then. Now, my life has completely changed.
First of all, it takes a while to build a new community. Grad school (and just life after college) is just different. You are no longer a part of the same social groups. You also have to find a project you are passionate about. For the first two years I was working on something I hated. I made a critical decision to switch and I haven't looked back.
Second, coming back to do PhD is not easy. As more time passes from undergrad, the more bills you have to pay, the harder it is to get recommendations, and the higher expectations will be. Passing your qualifying exam may be tougher.
Third, a PhD does not rail you to anywhere. It OPENS doors. American students do not fully comprehend this fact, and as such, American PhD students are the minority in most PhD programs.
American students are a minority in PhD programs because the risk/reward is traditionally awful, relative to the other options available to US citizens. The calculation is different for foreign students (particularly those from emerging economies), which is why you see more of them in graduate schools.
Mind you, this is not true for every US citizen, and it could change. If you can't get a job, getting a graduate degree is a lot better than taking to drink. (Unless your graduate degree drives you to drink, which I have seen happen. Try to remember that doppelbock is not a food group!)
But the original submitter is fortunate enough to have three job opportunities, at least, so that doesn't seem to apply in this case.
I would be interested in seeing how many PhD students come from affluent backgrounds. How do those statistics compare to foreign students and their backgrounds?
I think it becomes harder to do a PhD later in life, whereas it seems easier to me to do a startup at any time.
(Anecdotal evidence, obviously, and a small data set. You're welcome to bring your own anecdotes if you like, or even data.)
Second: I'm not advising that someone work in industry or on a startup for fifteen years and then go back to school. You can sample careers pretty quickly unless that career is "professor". I believe most startups do you the favor of giving you a quick but intense taste of the startup life and then promptly failing, saving you the embarrassment of having to quit. There's also no rule that says you need to sit at your desk at Microsoft until they kick you out: You can work there for a year or three and then go back to school, and the choice is yours. By contrast, it is really hard to leave grad school and then return, so once you're two years in to your Ph.D. program there will be strong incentives to stick around until you're done, even if it takes eight years. The only way out is through, unless you want to become what Matt Groening once called "The Bitterest Person in the World: The Grad-School Dropout".
Now, it is true that getting a professorship could take you a very long time, and that you therefore want to start as young as possible so that you still have some youthful vigor left to find a new career after you can't secure a tenured position. But you can spare a handful of years.
Ha!
I'll add the following quote, which I think is brilliant:
"Two years in [to a PhD program], and quitting will be like gnawing your own leg off."
(from: http://www.swarthmore.edu/SocSci/tburke1/gradschool.html, which is one of the better interweb answers to the "should I go to grad school?" question.)
"If you decide in your first year that it is not for you--indeed, suppose you conclude that you're better than all of this, a broader, richer thinker who can't be constrained by the ivory tower--you will still have to deal with the nagging fear that somehow, some way, you just weren't good enough, that you couldn't cut the mustard. That fear will almost certainly be wrong. Perseverance can get most students through graduate school. You should feel good about how well you know yourself if you decide to quit. But academia is a total culture. It changes your standards for what is good and what is bad, what is smart and what is dumb."
I don't think there's anything in his writing that suggests that you're less of a person if you don't go to grad school, or if you decide to drop out. It's just the culture that encourages you to think this way; you shift your conceptions of good and bad to fit the academic ideal.
Investing in a startup at least brings a chance for earning more money, whereas I would consider doing a PhD as mostly for fun.
Also, at least in Germany it becomes harder to get grants when you are older. For the common government grants there is an age limit of 30.
Many professors seemed to prefer younger student, too. I was about 31 when I started looking for professors to work with, and many said they prefer younger students.
I would say go for the startup idea since you see a need that could be better filled.
In fact, it's even worse than that: what you find interesting is not necessarily something that the average web surfer could even understand.
Of course, academia has its own problems (e.g., your peers don't really care much about your success and have their own agendas, which causes problems with peer review). Also, publishing stuff isn't really anywhere near as rewarding as having a successful web startup.
That being said, what do you really want to do? What are you good at?
I've written a bit about PhD life in another post (http://news.ycombinator.com/item?id=445082) and of course you are welcome to contact me if you are being pulled this way. Bottom line is that I think you should do the PhD if you are completely ready. Putting off a PhD will only make coming back to do it much more difficult, especially when bills start to pile on. Getting a PhD is not setting your life in stone, rather it will open up MANY MORE doors for you in the future.
The start-up sounds enticing. If you two agree about how much work needs to be done and this is an area that you are truly passionate about...just go for it! However, know that good ideas are not finite. There will always be time.
As for the job offers... Well, I tend to hate the corporate environment so I couldn't give you unbiased feedback. If you feel you can create value, and feel like your skills are valued, then perhaps its a good match.
With any of these, I would encourage you to talk with the people you would be working with. Your community will make a HUGE difference POST-college, because life after undergrad is shockingly different.
It would not make the PhD itself more difficult, but rather starting it - you are likely to have more and more responsabilities and burdens as time goes on.
However, if you have a family, a mortgage, a car, and a life outside of school that is time-consuming, getting the PhD will not be a top priority. This makes it difficult to do all the logistical things of getting a PhD too--it's not all about research. You have core classes to complete, many tests and defenses to pass, and conferences to attend. Although it's entirely possible for someone to do this, I cannot imagine it would be easy or good for your studies/other obligations.
Don't let money dictate your decisions. Seek to optimise your experiences and that goes for life outside of IT & work. The startup option will teach you the most. Just don't count on getting rich. It so spoils the experience.
Fundamentally, you will only know in retrospect whether or not you made the right decision. On the other hand, the stakes are not as large as they seem: 5 years is 125% of your time as an undergraduate, but only 12.5% of your career. (In other words, you have time to switch courses later, though it may appear now that you don't.)
Finally, the advice I always repeat to people facing difficult decisions. (Anyone who knows where this came from is welcome to tell me. I would attribute it if I knew the source.) .... When you have a difficult and important decision to make, flip a coin. Not because flipping a coin is a good way to make important decisions, but because in the moments while it is spinning in the air you will know which way you want it to land.
Trust yourself, good luck, and congratulations on earning such a plethora of attractive options.
The "Real World" is totally overrated. Entry level software jobs generally suck donkey balls, even if they sound like they won't.
However, it is very particular to your situation. If your job offer is to work on Photosynth at Microsoft Research, turn down the full ride to Toledo State U. However, if you're funded at Stanford or MIT, turn down the entry level code monkey job at Amazon or Yahoo!
Remember, you can always change your mind and any software company you join does NOT have to be a career - you can quit and start your own start-up at any time if you have the courage.
For one you may discover that you can do cooler stuff in less time and to more satisfaction. For two you may notice that not all PhDs are created equal.
I had a choice of doing PhD when I was graduating (that was in Eastern Europe and it required getting an approval, a recommendation and god knows what else, so it was more of a privilege than an opportunity) and decided not to. Now, looking back, I'm pretty sure the stuff I worked on in a past 10 years would easily get me a PhD or two. It also earned me enough money so I can actually go back to school and do that PhD ... only if I wanted to :)
2. PhD
3. Corporate job
Order is from "earlier the better" to "later is better" order.
I would also suggest you read this: http://www.amazon.com/Smart-Choices-Practical-Making-Decisio...
I can't imagine regretting punting a job offer from BigCo, but people are different.
The startup idea sounds cool, but making a better (or at least decent) mousetrap is really just an entry fee for attacking the sales/distribution problems associated with enterprise software. Hopefully you or your partner are good at sales and eager to do it.
I think most people are likelier to regret punting on a risky choice than the safe choice.
A PhD is lots of really hard and thankless work, and if you're not really excited about it at the beginning, you won't finish, and that's the worst thing you can do.
So - unless you're hiding your passion for research in your post, I recommend the startup.
I'm sure you know about them, but just in case: the AML field as represented in the Wikipedia article (and the sales report PDF it linked to) missed one company I know of who's been doing AML services 'since 1993', called Logica: http://www.logica.com/risk+and+compliance/350234053
A startup seems like a relatively quick return but its a gamble.
As for a career, both other options constitute as elements of one. If you mean striving to have some polished, dignified social timeline in your field, I say screw it, live for now.
Normally, based on my background I'd recommend the startup as #1, the job at Microsoft as #2 and the Ph.D last.
But I must say I'm not crazy about the startup you describe. Would I really advise you to join a startup I don't believe in?
Failure is a great learning experience. Looking for paying customers is a lesson that neither Microsoft nor a Ph.D will ever teach you. But isn't it a little bit early? My preference is to become an expert first and then start something. You must do your own soul searching. Are you satisfied that you are "great enough" in your field yet?
For some, it takes a few years of a regular day job. For others, they are great by the time they get their bachelor degree.
One problem will be getting the contacts. Also different banks have different systems. Maybe you can work something out with avaloq (their system is used more and more in different banks). Banks will never transmit data outside of their servers, and even people working there will not access any web service at all, if their computers are connected to customer data. So you have to sell the hardware as well. I think the biggest problem will be to integrate into the existing system.