October 2, 2014

How to do research

As noted in previous posts, I love reading, writing and doing research. However each time I consider return to academics world to pursue a PhD, my inner voice would keep yelling out loud I'm crazy. As normal graduate students, I know the agony of doing research. It's really hard to maintain enough energy to keep ourselves from burnouts. I was stressful and burned out once.

But recently after working in industry for a while, I realize that I didn't get around well because I didn't know how to conduct researches. I didn't know how to divide tasks into sub-tasks to conquer. I didn't know to that writing down ideas and keeping a research notebook is important. I didn't know how to set milestones and strive for them. when got stuck, I didn't bother to ask for helps from my professors or from friends. I, inversely, even thought that asking others show my weakness, which showing my arrogance. I didn't even try to visualize things in my brain, which I think I'm good at, and break down things to maintainable size. I learned by solely reading without trying to build a working toy product to see how it really works. My pictures of research objects are vague, unclear, sometimes misleading. I didn't know how to work by myself. I always ask myself why I still don't start a company and now I somehow figure out the answer by myself. It's because I don't have a clear goal and don't know how to do it by myself.

Rarely I can find a simple but outstanding piece of writing, but each time I do find out something it's really a masterpiece. https://people.cs.umass.edu/~emery/misc/how-to.pdf , a suggestion for graduate students of MIT AI lab is one of those masterpiece. It's a simple, clearly documented recipe for survive at AI lab. It posed a reality and it opened my eyes. Just bookmark here to in case I forget about it later.

No comments: