The CURF has exposed me to a new side of research where risk tolerance is high, collaborators are valuable, math is everywhere, and time is warped. I have taken classes on how to do research, but nothing prepares you better for research than doing it yourself. Several people have asked me countless times this semester, “What even is research? What do you do?” or “What are the main differences between your experience in industry and research?”. Personally, I love getting asked these questions because each answer I give, I learn a little more about what research truly is to me and how it is different (and similar) to a role in industry such as software engineering.
Software Engineering vs. Research
Over the past semester of doing research, it became clear to me how working in industry as a software engineer is different than doing research. For example, the risk tolerance as a software engineer may be much lower than a researcher because of the strict deadlines that could cost the company their reputation, quality, reliability, and money. Although, it isn’t necessarily the case that there is not any room for research in a software engineering role.
Researchers (in computer science especially) may have a much higher risk tolerance because it can lead to new discoveries. One of my favorite books, A Wrinkle in Time by Madeleine L’Engle, revolves around this idea of a tesseract which is a wrinkle in time. I analogize research to be a tesseract where you cross the boundaries of knowledge to find new ways to solve problems.
“A straight line is not the shortest distance between two points.”
~ A Wrinkle in Time by Madeleine L’Engle
Research and industry roles are similar in a sense that it is necessary to formalize a problem. In industry, there are clients or customers who can help define or formalize a problem. In research…you have to formalize the problem, convince others that it is an important problem to solve, and then solve it. In some cases, you have stakeholders to help you formalize a problem more. Throughout research, it was challenging to identify the best way to mathematically formalize my problem that would make sense to the machine learning community.
I was fortunate to collaborate with several other researchers and stakeholders at different points throughout my project with a wide range of backgrounds. I would undoubtedly say that their experience, knowledge, and feedback was a very helpful resource throughout the research process. With a project like mine, it is necessary to collaborate with researchers from other fields and interview stakeholders. It may be intimidating at first to think about reaching out to a professional in industry, but you will be surprised with how many people will answer your emails!
Dedicated Space & time
My CURF project stemmed from my previous research that I was supposed to conduct in Manizales, Colombia before the pandemic cancelled my trip. I never had the dedicated space and time necessary to progress this project to the point that we should present preliminary findings or submit a theoretical paper to a machine learning conference. The most valuable thing about my CURF experience was having support to dedicate the space and time needed to progress this project. Another valuable thing about the CURF is that I have been exposed to several other undergraduates conducting research in very different disciplines. Learning about their research and the research methods they use has contributed to my overall understanding of research.
Where I’m Going Next
I plan to finish this research project by the end of May to submit our findings to an international machine learning conference. Aside from completing this project, I am super excited to announce that I will be staying in Pittsburgh to complete my PhD in Human-Computer Interaction at Carnegie Mellon University. I am planning to focus on incorporating fairness and mitigating discrimination in algorithms used for high-stakes decision. I hope to apply the skills I gained throughout my CURF project to projects I will work on during my PhD.