Why I left the tech industry for grad school
“Why are you leaving your job?”
“Don’t you make enough money?”
“Why would you leave a big tech company?”
“Are you going for a bigger paycheck? You must be stuck at the same level.”
“What?! You said you’re completely leaving the tech industry?”
“Oh! PhD? Don’t you already do research at work? Everybody learns at their jobs, just hang in there and you will too!”
These are only a few of the questions and comments I received after delivering the jaw-dropping news that I was leaving my job to go back to school. Having only been in industry for four years when I decided to quit and go back to grad school, I did not feel too attached to the comforts of a 9-to-5 job at a large corporation. But for a lot of people, the notion of a stable job and income is very important, and turning it down to go back to school, especially for a long-term commitment like a PhD, is hard to fathom. I completely understand that, and make no mistake, all the perks of an industry job like a having a proper routine where you can say no to work after a certain hour, expecting to get paid a decent amount on a certain day every month, being able to progress up the corporate ladder, and always having a clear trajectory in front of you are indeed hard to say no to (also, free sodas!). I am also aware that it is an immense privilege to be able to transition from a high paying job back to school and put your own aspirations before the needs of those who may rely on you.
So why did I decide to go back to school? I have tried to rationalize this more times than I can count and every single time I have come out unconvinced by my own reasoning, yet inexplicably certain of my choice. Typically, when people make a similar decision, like switching jobs or companies, or selecting one job offer over another, they have a calculated list of pros and cons that makes their decision seem very obvious. I did the same when I was finishing up my master’s and deciding where to head next. But my choice to pursue a PhD was less black and white. I was already doing research in one of the leading industry research labs in my field at Microsoft. I was already publishing papers and learning from other leading researchers. I already had a fair amount of autonomy in deciding what I wanted to work on and, in the process, how I wanted to approach those problems. Heck, I was even a mentor to several PhD students over the years (who are now much senior in their PhD journey to me and will be my mentors—how the tables have turned!).
Despite all of this, I had an inkling that I should be back in school. While Microsoft was a great place to work and transform research ideas into products, I wanted to receive formal training in how to properly do research in a more autonomous, academic setting. In the corporate world your performance evaluation and career growth are mostly a function of what you have concretely delivered and it is implicitly a portfolio of your successful endeavors. This, inadvertently and unconsciously, filters out a lot of ideas and potential avenues for exploratory research. In my case, for example, I was frequently thinking about the feasibility and practical aspects of deploying my research ideas into Microsoft’s cloud platform. Having that perspective along with a plethora of real-world data from large scale cloud systems was imperative in targeting critical problems in production and creating meaningful impact for customers (e.g., better resource utilization of machines in the datacenters means lower operational costs for the company which eventually reduces the price of cloud computing for users). However, at times I also thought about the value of doing science for the sake of exploring without being tied to a particular agenda or improving a particular metric. Maybe I was glamorizing those tales of discoveries through complete serendipity: Penicillin in a petri dish, x-rays for imaging, and radioactivity from Uranium. Maybe I wanted to step into a world with more unknowns and explore topics that interest me without the guilt of not making progress at my job. Whatever the reasons may be, for me the journey to becoming an independent researcher, facing the ups and downs of research, taking ownership of the process and the final outcomes, and just the experience of being in academia was more important than the steady income and stability offered by a corporate career. There are of course tremendous knowledge gaps that are much harder to fill outside of school. For me the joy that comes knowing that I can learn from some of the world experts in an open-ended way and apply the learnings in my own research is very exciting.
When I heard back from MIT, it was hard to believe that I officially did have an option to pursue something I had never completely eliminated from my subconscious despite me always telling myself that only crazy people commit to something so long and time consuming like a PhD. Maybe it was the thrill of the challenge? Maybe it was the yearning to be back in school? Maybe it was the aspiration to one day join the other side and be a professor? I’m still not entirely sure what my motivations to complete this PhD are but I am very excited to find out as the journey unrolls.
It hasn’t been long since I’ve started grad school but I can already feel a change in my stride as a researcher as I experience the contrast between working in academia versus in industry. For example, we are building a system for improving the performance of applications like training machine learning models and data processing engines while also lowering their operating costs (e.g., computational resources and energy). There are too many open ended questions, reliance on pieces of hardware that are not commercially available, and the project has a high risk of not working out the way we want it to. But being at MIT, we may get access to some of the necessary hardware to build our own prototypes before others do. We may have to rope in more people who are experts in other related fields, but we can find them here at MIT. We can keep pursuing a project until we succeed or can concretely write about why it may not work, and both those outcomes are completely fine.
I will report back on how accurate my preconceived notions of doing research in academia were. From what I can tell so far, this is going to be the journey of a lifetime and I am looking forward to every step of it.
Share this post: