Humans of SymSys: Emily Hu
/
Who are you in a nutshell?
I'm a recent graduate of Stanford Computer Science (2020, the year of Zoom Graduation!) and a current coterminal Master's student in Symbolic Systems. My focus for both degrees is human-computer interaction, and I am broadly interested in using technology as a lens to understand human behavior. How does technology change the way we communicate and organize ourselves, and, in turn, how do we use it as a lens to view these changes?
I've spent some time as a product manager in industry (I was a Google APM this past summer), and I'm also a TA at Stanford. I most recently staffed Minds and Machines (Fall 2020) and Social Computing (Spring 2020)! Teaching virtual sections has been unexpectedly fun. Pre-pandemic, I was heavily involved in the social dancing community, and co-directed the latest Stanford Viennese Ball (back in February 2020--- what turned out to be the last weekend of viable social activity). Dancing has translated somewhat poorly into our quarantine-world, but I certainly dance on my own, and according to Spotify, have listened to over 60,000 minutes of music this year. Outside of all of that, I love to doodle and have picked up calligraphy. I occasionally cook things and enjoy making (and drinking) both coffee and tea.
Why SymSys?
I think I'll have a different answer to this than most people, and that's because I was a Computer Science major (undergrad), but I'm a SymSys coterm (Master's). Whenever I tell people this, I feel like I surprise them a little, because most people do it the other way around --- SymSys undergrad, Computer Science coterm. However, I wouldn't change anything about my path; I loved being part of the Computer Science department, and doing a "deep dive" on the technology aspect gave me a solid foundation from which to branch out and understand the human-centered implications that I'm now focused on.
I chose the SymSys coterm, as opposed to the CS coterm, for a few reasons:
It's a research-based degree program. In college, I unexpectedly discovered research as a genuine passion. Whereas a CS coterm would require me to take more classes, a SymSys coterm is designed around the research, and ends in a capstone thesis. This doesn't mean I don't get to take awesome classes (I do!) but it means that I also get to tie it all together with a project that I love, and that I get to call my own.
It's more flexible and interdisciplinary. You can see from my introduction of myself that, while my work is rooted in my computer science background, I also bring in elements of psychology and social science. I study humans and technology as interlocking systems---and what better way to do so than via Symbolic *Systems*?
The community is smaller --- and therefore closer. Each quarter during the coterm, we have biweekly meetings with all the Master's students, and we get to hear about each other's research progress. You genuinely feel supported; that personal touch is really important to me.
What is your concentration and why?
Human-computer interaction. I was a CS major for undergrad, but the SymSys HCI program is fairly analogous. Both of my best friends and close study buddies/collaborators were SymSys HCI, and we shared almost all of our classes in undergrad. The main difference, I think, was that I took a few more computer science / engineering classes in the end, and they did a little more philosophy/linguistics.
My first exposure to HCI was very early on in my career, where I did the CS Undergraduate Research Internship (CURIS) and joined a project in the HCI lab. I spent the whole summer listening to talks about research from all the different concentrations (AI, Systems, Theory, etc.), and HCI just seemed to resonate with me the most. For me, studying computation was interesting because it gave me a toolkit with which to solve human problems. And that was really the initial reasoning for me --- I wanted to major in something that allowed me to impact humans.
Over the years, as I took more classes and started carving out my own niche, I've only been more convinced that HCI was the right choice.
What’s your favorite SymSys-related class that you’ve taken?
I have two -- CS 247 and CS 124.
I loved CS 247. Human-computer interaction classes are so fun, because they're very hands-on---focused on working in groups and practical applications. When I took 247, they broke the quarter into a few smaller projects, and I had a fantastic time learning about all the design and prototyping techniques in each one. This is probably the single class I draw on the most in my everyday work, whether it's industry or research, because knowing the interaction design process is essential everywhere you go.
I also loved CS 124. This is probably the most "SymSys-y" class I've taken, since it's quite literally a combination of computation and linguistics (to study NLP), and in the end, we made a bot (which brought in a little bit of psychology, especially when we discussed early versions of mental health-focused bots). I had so much fun in this class. When we made the bot, we basically went wild with extensions, and it was just so fun to be creative and hands-on.
Are you involved in research? If so, what are you working on?
Yes! I mentioned this before, but I study the intersection of technology and humans---specifically, organizations, groups, and teams. My undergraduate honors thesis was about the decision-making of online groups. I wanted to know whether groups make consistent decisions together; so, I designed an online jury study, where we had participants decide one pair of analogous scenarios by themselves, and another pair while participating in a jury. This was a within-subjects study, since each participant experienced both conditions (in a random and counterbalanced order).
We then measured whether groups decided more consistently than individuals. To my (and my collaborators') surprise, groups and individuals were equally consistent. Also interesting was the fact that aggregating individual decisions into a non-deliberative "vote" led to decisions that were nearly random. We found that deliberation was able to secure a stronger winning majority, whereas voting on your own meant that one or two swing voters could completely flip the result.
A project that I'm working on more recently is understanding how people transition from collaborating in-person to collaborating remotely. Here, I'm doing a field study of a real organization, and trying to understand what you need to change about your practices and decision-making processes in order to work as a distributed team.
What is one piece of advice you'd like to offer to younger students?
It's never too late to start something, and it's never too late to stop something.
I never thought research was going to be something I could do, because I hadn't done it in high school---I knew people who had already won multiple science fair competitions by the time they were 18. I was also worried about studying computer science in college, because my only coding experience prior to Stanford was using Macros in Microsoft Excel (LOL) and listening to the first ~4 lectures of Harvard CS50 on edX.
Well, it turns out that it totally wasn't too late, because I'm here today. I often talk to younger students who say that they're worried about trying computer science, or about applying to internships in software engineering and PM because it's "too late to start." I think it's never too late. I promise, there is a place for you here.
Conversely, I think it's never too late to quit doing something, if you realize it's no longer something that you're continuing to learn and benefit from. For the first three years of college, I was an active competitive debater. But over time, I realized that I was no longer pushing myself to learn and grow through debate as I used to, and I had to learn to let something go, despite all the joy that it once brought me.
So, the lesson that I'd tell younger students is that it's never too late to learn something new, and it's also not too late to reflect on something you've done for a long time, and choose to redefine yourself.