by Rohit Agarwal (PhD Fellow, UiT Tromsø)
Self-question: Why research?
Answer: I like to read research papers. I like thinking about possible problems and comprehending their solutions. I also like to write papers because they give me a sense of achievement. Hence I want to do research. Before deciding on pursuing research, I did work in industry for a good 10 months. I did not hate the job but I was not excited either. I did not feel any sense of achievement there. Though my job was in the field of cloud computing and not in ML, I feel my experience would have been the same.
“ The aim of a PhD's to ensure that no one, including your advisor, understands what you're doing after the first couple of years.” ― China Miéville
Self-question: Why then a PhD and not a job in industry R&D?
Answer: The biggest issue I think right now is that no good R&D hires a graduate student (this is my thinking, I might be wrong). Each company wants to hire a PhD (for ML field), so it becomes imperative to do a PhD. I think this is fair from an industry point of view because there are a lot of concepts, material in ML. And a PhD gives an individual to go through them, form their understanding and contribute to the field. Having said this, I think even if I would get an offer from an industry R&D at this point of time, I would rather choose a PhD in academia. Reason for this is that I have more freedom on the topics I want to research and I will be accountable for it.
Self-question: Why machine learning?
Answer: I like mathematics. Solving equations and making one is what excites me.
Self-question: But then there are a lot of fields other than ML that use maths. So, why ML?
Answer: Well the simple answer to this can be the peers and current research scenario. When I was doing my bachelors, every cool stuff that I heard about was related to ML/AI, be it in academia or industry. So, I got attracted towards it. I started learning out of fear (worrying about my career) and peer pressure (self-thought), but gradually I started liking it. The application of maths in solving practical problems excited me. I did three internships in ML during my bachelors and each one had a different impact on me. My experience improved internship by internship. Even now, I have doubts sometimes about doing a PhD, but I am 100% sure that this is best for me at least for the next 3–4 years.
Self-question: What would you want to do after your PhD?
Answer: I don’t know for certain and I don’t want to think right now. I can continue in academia, go to industry R&D, go for a data science or software job, or become an actuarial scientist. Who knows? The way I think now will change in the future. I will decide then based on the situation and what I feel is right.
Self-question: Do you think you will regret your decision to do a PhD?
Answer: Of course. I am 100% sure, I will regret it sometime. I have this philosophy that no matter what you do in life, you are going to regret it sometime in future. But, the only thing we can do is decrease the regret-span. What I mean by regret-span is the amount of time one wastes regretting about their decision. Become rich and you will in future miss your early days which will turn into regret (albeit for a shorter regret-span). So, yes I know I will regret my decision, but I will try everything to make this regret-span as little as possible.
P.S.: These are all the questions I ponder on whenever I go to sleep. So, I thought of writing it down along with the responses I come with. I think this gives me a clarity of why I am doing a PhD. These are my thoughts and I understand that others may disagree with my thinking. I would love to hear other people’s thoughts on why they are doing or want to do a PhD. If there are any other questions you may have for me, please ask.
Follow Rohit Agarwal at Medium (Click here)