Jayant Haritsa (IISc Prof) on MS vs Phd
My last post on MS vs Phd got a lot of responses and varying view points.. So, I thought I would get view points from different people. Today, I have Jayant Haritsa – Professor, SERC and CSA department of Indian Institute of Science, to share his views on MS vs Phd. Thank you very much sir for sparing your time and answering me over email.
Jayant Haritsa’s views on MS Vs Phd.
A couple of random points (caveat: as you will see below, i am heavily biased towards a traditional notion of academics):
1) Firstly, we should not make value judgements between Masters vs PhD in the sense of one being superior or inferior — it is more an issue of what *you* would like to do in the remainder of your technical life.
If you are the type who prefers to work in teams and is keen on developing immediately useable systems, then pursuing a Masters would make good sense. On the other hand, if you are the type who likes to work alone and secretly wishes to pursue an irresponsible life and be paid for it, then doing a PhD may be an attractive proposition.
2) I personally am not in favor of PhD thesis coming out of large systems projects, based on the following reasoning - a PhD should gear you up for carrying out *independent* research and is in a sense, a process of *self*-realization. However, often in a large group project, you get either lifted or pulled down or channeled by the group objectives and members. So you really don’t get to figure out your *own* abilities due to the external modulation. Ideally, a Phd should prepare you for the loneliness of academic life
Further, in large systems projects, sooner or later, getting the software out usually becomes more important than the concepts themselves – but to me the quality of a PhD is defined by the fresh *conceptual* breakthroughs.
3) The reason often given for an industry stint before or during PhD is that you get to grips with “real problems”. My view is again that of the pure mathematician – who cares? The “real” just means that it may have *short-term* applicability, but in practice, most PhD thesis find their way into the real-world in the long-term. To give you a recent example – Mike Stonebraker is now using some of the scheduling ideas from a thesis of fifteen years ago in the Streambase prototype.
Do a Phd when you are at the peak of your mental powers, which is *right now* - don’t bother so much about whether it is “industry-relevant”, as much as “did I have fun thinking about the problem and did I come up with elegant solutions that I am proud to show my mother”.
4) Finally, in my view, ideal Phd is one where *you* come up with the problem, work alone (modulo advisor) and single-handedly write a definitive thesis on the issue.
Thank you sir! Readers, Hope these inputs were helpful!If you are in NIT Trichy, there is a resource on apping tips, MS vs Phd etc in TnP noticeboard! Dont miss out on that if you are planning to app! The doc was compiled by Vicky and GREram – students of 2006 batch, now pursuing phds. I am also getting the views of MS vs Phd from a couple of other sources. A Phd student in Stanford, an MS student, HR department on how they view MS vs Phd in terms of jobs etc.
Hope you enjoyed this episode of Coffee With Experts! Previous episodes of Coffee with Experts is here.
First Posted on Coffee With Sundar – http://coffeewithsundar.com