Doing a Machine Learning PhD While Working in Japan

I recently completed my PhD in Computer Science at Tokyo Institute of Technology (now the Institute of Science Tokyo) as an American student on the MEXT Scholarship. Simultaneously, I continued my work as a software engineer and researcher at Google. Because it’s relatively rare for an American student to complete their PhD program abroad—as compared to the number of international students that study in America—many people eventually asked me, “But why would you choose Japan?” This comes with the implication that, given the American tertiary education system’s prestige, I was being foolish by not taking advantage of it. This point of view has merit. American schools do have incredible name recognition worldwide, as well as opportunities that are rare elsewhere. But I chose Japan, and would like to walk you through my experience to give you the pros and cons. I admittedly did not have a typical background going into my PhD: I had a good fallback since I had already been working at a FAANG, I had meaningful savings, I already had research experience, etc. This made my time in Japan significantly easier and less stressful than for many of my peers. But it was still not a walk in the park. Doing a PhD is difficult and isolating, which is exacerbated by being an international student in an unfamiliar country. And so, while I overall very much enjoyed my time there and have freely recommended it to many people, the decision to pursue a PhD in Japan should always be considered carefully. Background Why Japan can be a good place to do a PhD Why Japan can be a bad place to do a PhD Finding the right lab Funding the degree: MEXT and the three-year clock Working part time during a PhD Social life at the university Living in Tokyo After graduating Epilogue Background I’m an American and did my undergraduate degree in America, and before Japan, I was a PhD student in Korea. I didn’t complete my doctorate there, but left with a master’s to join Google full time after an internship. My original research area was automata theory, but at Google I work primarily on on-device language models for keyboards. I have always inhabited an area between pure research and engineering and worked hard to maintain that balance. I knew that I wanted to finish a PhD eventually, but I had a fairly specific constraint: if possible, I wanted to keep working part time while doing it. I knew there was some precedent for this arrangement at Google, but it was rare. At the time I started looking seriously at doctorate programs again, I was working from my parents’ house during COVID-19. I had gone from Seoul, to New York for an internship, to Mountain View for a full-time role, and then back to my hometown in the American South. I had enjoyed some of those moves more than others, but the overall effect was that I had a pretty good sense of what I liked and disliked in a place to live. I loved dense cities, I liked being abroad, and I knew that if I was going to do another PhD, I wanted the location itself to be a meaningful part of the experience. Why I considered Japan When I started looking into PhD programs again, I was not particularly looking for something in Japan.