Intern Success Stories
In the summer of 2013, I did an internship at MSRC, working with Milan Vojnovic and Christos Gkantsidis. After some years, I might forget which pub is my favorite, which beer will knock me out, or how long the day time is in the summer of Cambridge. However, I know the numerous discussions with my mentors will stay in my mind for a long time: those discussions not only resulted in the joy of progressing, but also brought me precious lessons.
If we present our ideas to an audience in an effective way, we will receive valuable feedback from them. In the first half of my internship, I worked on problem definition, and the goal of our discussions was to evaluate the appeal of the formulated problem. In many discussions, I talked with Milan about related work and possible theoretical results that we could deliver. I did some confusing presentations in our early discussions: spamming symbols without serious definition and making whiteboard notes like doodles. Fortunately, I had a patient and kind audience who pointed out my presentation issues and gave me advice to make them clearer. Those suggestions indeed helped: I could keep my audience engaged even in a two-hour long discussion. Moreover, I started receiving valuable comments, including the fault in my proof, new ideas to analyze the results, and new directions to explore.
Christos once told me "In most cases, we don't know if we are going to fail; however, we can see if we fail fast". After a problem is defined, it is time to find out solutions. We worked on a hard problem, and there was no clue of how to deliver good solutions. At that moment, we had some intuitive solutions, however, we did not know how well they would perform and were struggling to find any theoretical support. In short, we were stuck, and I found Christos to share the pain. In the discussions, Christos encouraged me to perform empirical study: as long as we could quickly see that any proposed solutions did not work, we’d have time to switch directions; moreover, he shared suggestions on experiment design and result analysis. It is the empirical results that inspired us to deliver good solutions to that problem afterwards.