Why should I do a Ph.D?

Question

Long back Bikram Ballav suggested me to write about my Ph.D. experience which I'm yet to complete. But after seeing the information level of the GATE takers I'm forced to share everything I know about Ph.D. and clear many misconceptions. So, why should one do a Ph.D.?

1. Because I'm mad - yes, those who do Ph.D. are mad people.
2. No other option. I did not get a good job.
3. I want to become a Professor.

Not sure of any more reasons but these are the common ones. Now, what one does during a Ph.D.?

1. Do TA work
2. Read 1000s of research papers and behave like Einstein
3. Publish papers to journals and do no coding
4. Struggle to get a gf/bf as no one wants a Ph.D.
5. Struggle for money while your friends earn plenty

Now, after having seen Ph.D. life in IISc., India, INRIA France, and OSU United States, I can say something.

1. Ph.D life is so different at good places like IITs compared to other Universities.
2. As a Ph.D. one is expected to work on one or more problems in some specific area. So, after finishing Ph.D. you become an expert in that field and any company working in that field should be happy to hire you. Yes, Ph.Ds from good places are hired at very good salary by Tech. companies.
3. Ph.Ds are not expected to code. But based on your project you will be required to implement something and that requires coding. During my Ph.D, I had done less coding than my Masters as for research more time is spend for analyzing stuffs.
4. Me and many of my batch mates got married during Ph.D. But this might be hard in India or US as Ph.D. life is more hectic there. 
5. Ph.Ds are paid okay. In India it pays better than a starting IT job whereas in France and US, it pays way better. One advantage of Ph.D. abroad is for 2 years there is no need to pay income tax.
6. Ph.Ds are not publishing bogus papers. That happens at low tier universities. At good places they struggle to get paper published at top notch conferences where the papers are reviewed by experts from the best Universities, industries like Google, Intel, Microsoft etc. And that is the value of top conferences, citations etc. No good Professor/researcher will like to have his/her name on a bogus paper.

Now, can I do Ph.D.?

I want to do what someone tells and lead a happy life earning salary they give

I want to do what I like and give my contribution to the world however small that might be

If you think like first- no, you should never consider Ph.D. If you think like second yes, you can do Ph.D. There is no super skills required for this. You get enough time, but you should spend it in useful ways. Anyone who has prepared by self and got to top 1000 rank is qualified to do a Ph.D. - because it requires self work and guidance as Professor can only guide you, since you are doing a new work he can only solve troubles you have. Most of the problem solving path, must be found by the candidate himself. Those who ask for questions like "what is an automata" are not fit for Ph.D. As a Ph.D. candidate you are not expected to know everything -- you can happily say "I do not know to anything you do not know", but you must be willing to learn anything that is required.

Who are preferred by IISc/IIT professors?

Most of them know about Indian education system. So, never try to fool them. They want to take those students who will do good work under them. Some prefer disciplined students who put lot of effort in their work - usually who tops university exams, some prefer those who have good basics - who can answer GATE level questions, some prefer those who have in depth knowledge in their areas of research etc. Also, those who have done some good work either in terms of publications, or coding projects are also preferred.

Comments

Awesome post @Arjun sir, yesterday I asked you about whether one should pursue Ph.D. or not and now many things got cleared, so thank you very much for writing this but can you please tell me the relations between publishing some papers vs real life implementation i.e does those papers helps in making the world a better place? the only example I have is Larry page and Sergey Brin, they were pursuing Ph.D. and their thesis later became Google company so they actually converted those research into real life implementation, and this is pretty rare so what is the use of those papers if it doesn't contribute to our society and won't change the world? Maybe I'm wrong and I'm not aware of some aspects but these are some questions comes in my mind when I think about Ph.D., the better way for me to contribute towards society is to create a company which will change the world instead of publishing few papers that no one will read or will just get some good rating in journals and no one will implement it in real life? Please reply sir if possible and thank you very much for everything you're doing for Gate aspirants!

---

I'm considering only research at good places.

What percentage of research publications translate to use products? I'm not sure, but certainly not more than 50%. This is because research is about problem solving and many problems might not have solutions or might not have solutions which can be implemented practically. The whole purpose of research is to propose ideas which can be implemented or to suggest features which are to be provided (by hardware manufacturers) for efficient solving of problems. Most times researchers do not implement stuffs but in certain areas that can also happen where Ph.D. graduates directly starting their own companies (this is encouraged in Europe). But imagine someone working in architecture. He can only suggest architecture features in his thesis. Big companies like Intel then decides if those proposals are worth being implemented. It is also true that while writing a research paper we might miss many practical scenarios but if a paper is accepted at the best conferences it means most of the practical cases would have been covered in the paper. Also, most good papers have enough readers because citation count is important for researchers and they publish papers at places where others read them. Those who work at big companies also read these papers and try to get implementation ideas for their new products.

---

+thanks Sir.

---

Thank you very much, sir.

---

@Arjun sir, how is doing MS in CS after Mtech from non iitian clgs like iiit/middle nits ?

---

Arjun sir, I have got a direct PhD offer from IISc at the CNS department. Ever since I was in class 9, I got interested in neuroscience when I heard a lecture by Prof. VS Ramachandran, the world-famous Indian- American neuroscientist, and my love for astrophysics was always with me since class 6. Somehow I passed my +2 with a decent rank in a state-govt college of my home state. For the first time in my life, I was formally introduced to computer science, prior to that, my knowledge of computers was limited to word excel, and games which by the way most kids had at my age.

Now, with the advent of AI, and with the sudden boom of DL,I am kinda interested to do research ( or at least try ) at the intersection of neuroscience and AI, what do you think what would be the career prospects after having completed Ph.D. from IISc.

I am ready for any research roles outside India ( I don't have any homesickness) , so if there is an opportunity outside India,I am happy to grab that as well ( Anything which is the at the intersection of Computer Science, physics, and neuroscience, I am ready to take endeavor to explore more on that !! ))