Making tech more user-friendly: PhD alum Chung-Hay Luk researches how people interact with technology

January 29, 2019

“Innately, I’m interested in trying to understand people and how they think and what their preferences are.”

Chung-Hay Luk, PhD program alum (entering class of 2005)

Chung-Hay Luk has two lifelong passions: a fascination with how the mind works, and a love of making things with electronics. As an undergraduate Bioengineering major at UC Berkeley, Luk helped develop an electronic nose in Noam Sobel’s lab, and then went on to earn her PhD in the Helen Wills Neuroscience Institute, studying decision-making in Joni Wallis’ lab. While in graduate school and afterwards, she engaged in a number of side projects that allowed her to creatively use her “tinkering” side, including making a giant, light-up, interactive neuron, and a dress with embedded electronics that could change colors based on votes submitted via a web app.

Luk’s combination of interests led her to do technology research and design work after she graduated from Berkeley, starting at a brain training company where she worked on games aimed at treating patients with addiction. Here, she became interested in user experience (UX) research so she could develop technology that patients would be motivated to use. Luk started her UX research career at a company that uses Google Glass to help doctors take notes, and she is currently a UX researcher at Google, working on their Photos product. She also does UX consulting for a startup company.

Read the following Q&A with Luk to learn how she transitioned from neuroscientist to UX researcher, her “After Science Science” experiments in grad school making origami out of dried kombucha, and how her current job allows her to do research wherever she goes. This interview has been edited for brevity.

Rachel Henderson: Tell me how you first got interested in neuroscience.

Chung-Hay Luk: I didn’t have a very clear path. For grad school I would have been in bioengineering for some schools, but for other schools I would have been in neuroscience. I was interested in the intersection between neuroscience and the making side of things. It’s a little bit more engineering, but not necessarily about hardcore engineering in terms of things like making prosthetics.

I considered a range of labs when I decided to choose Berkeley. I was thinking maybe Shaowen Bao’s lab or Joni Wallis’ lab or Jose Carmena’s lab. I’ve always been interested in how thoughts are created, so that dimension of neuroscience was interesting for me.

As an undergrad, I was in bioengineering at Berkeley so I didn’t move very far, I only moved from one side of campus to the other side (laughs). I was already leaning more towards neuroscience in my junior and senior years when I was in Noam Sobel’s lab, studying olfaction.

Photo: A model (center) wearing a dress Luk (left) designed that changes color based on votes submitted via web app at the 360Fashion Show in 2015. Photo by Shenova.

RH: What did you do in the Sobel lab?

CL: I worked on a couple of projects. One was an electronic nose device that was a partnership with an industry partner. I was soldering the circuit board so that we could test the electronic nose with different odorants and see if the signal would be different for the different odorants, so we could then predict what the nose was smelling. For the project that ended up being my undergrad thesis, I was trying to figure out: can you parameterize different odors based off of how they smell perceptually as well as how fast you are at differentiating different odors?

RH: You stayed at Berkeley to do your graduate work—did you look at other schools?

CL: I did look at other schools, definitely. Personally [Berkeley] just ended up being a good place, given my broad interests. I didn’t have a very clear direction, it wasn’t like I wanted to study “x”, so I couldn’t really narrow it down to: “These are the top researchers, I should be going into their lab”. I was looking for a place that was strong both in the more computational side of things, as well as the biological side of things.

RH: Why neuroscience versus bioengineering for your graduate program?

CL: I did bioengineering at Berkeley as my undergrad, so I wanted to learn more skills and the Neuroscience Institute was a good place. For instance it’s smaller—it’s not like the [undergraduate] Berkeley experience where you’re one small person in a really big pool. I really liked the intimacy of this very small program. I had a lot of resources there. They’re a very good interdisciplinary program because they have professors across a lot of different disciplines. So that way, if I wanted to learn anything it seemed very easy to approach different professors or different grad students in different labs to learn from each other. I liked that part.

Also, I learned that engineering and science come at problems from very different angles. I didn’t fully appreciate that until I was in grad school in science and going, oh, okay, people think about different problems or value different techniques, and just in a different light than I’m used to.

Photo: Luk (center) catching up with HWNI PhD Program alums Liberty Hamilton (far left) and Jen Sloan (far right), and former Wallis lab members Alejandra Dominguez (center left) and Erin Rich (center right)

RH: How did you choose your thesis lab?

CL: It was a really hard choice, there were a lot of good options. I ended up choosing Joni Wallis’ lab. It was a good cultural fit, as well as [research] interest. In terms of culture, I’m more of an introvert, so I wanted a lab that was a little bit on the smaller side so I could have time to explore and kind of create my own space. Also, whenever I needed questions answered, I could just quickly get answers.

Joni Wallis is also really, really strong in statistics, which was a specific skill I wanted from grad school because I knew that that was a very good transferable skill and it would go a long ways. Then the other part was the topic matter. Decision-making ended up being really interesting for me. I think innately I’m interested in trying to understand people and how they think and what their preferences are. So joining the Wallis lab was very good for looking at that from the foundational level.

RH: You mentioned considering statistics a good transferable skill. When you started grad school, did you think you wanted to be an academic scientist or did you want to go into industry, or were you open to any options?

CL: So my dad’s a professor (laughs) and I didn’t want his path exactly. But I did a lot of research, I’ve always been kind of that inquisitive kid that touches things and pokes around. I knew that I wanted to do something research-y. I was planning to go into industry—I was thinking I’d go into an R&D department at some company. I didn’t plan to be a professor. I don’t know what will happen in my future, but it’s not something I’ve pursued.

RH: Tell me about your PhD thesis project.

CL: I was looking at the difference between stimulus-outcome and response-outcome encoding in the prefrontal cortex. So that’s kind of a psych perspective, as in habit formation. If you have a stimulus that triggers a response, there’s an outcome, and how do those paired associations get encoded in the brain and what are the differences?

I’ve realized since I’ve moved into industry, that kind of very academic, psychological framing of my thesis doesn’t really resonate, so I can spin it in different ways. If it’s more of a computer science crowd, I will talk about it more as a credit assignment problem—how do you figure out what the outcome really was caused by? Now that I’m in user experience, I’ve actually reframed my thesis, so I talk about it as passive viewing versus active engagement on choice behavior, so it has words and language that the audience can relate to.

RH: Was user experience (UX) on your radar as something you were interested in doing as a graduate student?

CL: I don’t think UX was very developed as an industry when I was a grad student, so it wasn’t necessarily on my radar. I ended up pursuing UX as one of the possible avenues. I also looked at more R&D style jobs and also looked into consulting and realized that it was probably not for me. I also considered technical consulting, like Exponent, and realized that you might actually have to testify in court, which is a scenario where I would feel very uncomfortable (laughs). I feel like I expanded my horizons, looked at a lot of options. I also had a crazy Google doc that had a lot of different neuroscience-y companies or startups to see what I might do. I’m happy to share that, although it’s a little bit old now.

RH: How was your experience at HWNI in general?

CL: It was a really, really good experience. I am still really good friends with many [of my fellow graduate students]—we still hang out. It’s also really great to have a sounding board, because sometimes when you’re in industry it’s really hard to figure out what you should be doing, what career path to take. There’s always a cohort I can lean back on. We also had a lot of fun, we traveled together, and we did side projects together.

RH: In terms of side projects, I saw that you collaborated to create a “NeuroDisco” with HWNI alum Erica Warp. Tell me about that.

CL: We made this over six-foot-tall neuron, which originally was supposed to be a science exhibit at the Cal Academy of Sciences for kids. We made it to show how a neuron works, as in, if there’s a neurotransmitter, it gets attached to a neuron and it’ll trigger an action potential. We made it so that it was interactive, there would be a little ball that would get slotted in, and the neuron would light up with LEDs [illustrating] an action potential. Erica Warp’s husband [Richard Warp] does sound, so it would actually make a sound that kind of sounded “electric”.

RH: Did you do that as a grad student or was that afterwards?
CL: That was afterwards. During grad school, I did other projects. For instance, with kombucha, you can harvest the scoby—the layer that’s yeast and bacteria—and if you dry it, you get something that’s like a paper-meets-leather kind of material. So I grew that on top of my desk. (laughs)  Me and a couple of other grad students did “After Science Science”, playing around with that to see if we could embed electronics into it. It’s not very easy, it turns out, but it makes for a very interesting material and we managed to fold origami with it!

Photograph: Luk and Richard Warp working on the lighting and sound components for the giant neuron.  By Erica Warp. 

RH: What did you do after you graduated from Berkeley?

CL: Afterwards, I went over to one of Michael Merzenich’s companies, Posit Science. They’re based in San Francisco and they focus on brain training. I was interested in that because it was a nice kind of transition. I still could use that neuroscience piece and I could start applying it in the real world. Over there, I was focused on exploring the idea of brain training—software programs that hopefully medical doctors would prescribe to help cognitive disabilities or conditions—for addiction.

So if addicts were to get brain training, what kind of training might help? What would it look like? It was really cool because I got to work with actual clinicians and doctors to figure out what kind of games would work. I would also be reading literature to see, based off of studies, these are maybe the active ingredients for what the games would need, and test them out.

RH: What did you do after Posit Science?

CL: [At Posit Science,] I liked a lot of the pieces where I was doing usability testing, or just trying to figure out how to make software good. If you’re going to prescribe software as medication, you need your patients to actually do the exercises—use the software, basically. So that starts touching into the user experience side of things. I really liked that part.

I ended up moving over to a startup called Augmedix. They use Google Glass to take notes for the doctor. It’s like having a remote personal assistant. They were looking to take on a user experience person, the very first one, so they were willing to take a bet on me. I was literally doing a career transition at that point, because I was going from neuroscience into UX. They were willing to take a gamble on me, which was awesome. I ended up doing both user experience research as well as design. The research side was trying to figure out if my designs actually were working, do people understand it, all those details. Then I would deliver the design—so what the engineers should be building.

RH: Since this was a new field for you, how did you learn about it?

CL: My planning was, let’s see how much I can do on my own. I figured, I have a lot schooling, it’s not directly applicable, but maybe it’s good enough and I can just kind of bootstrap the little pieces I’m missing. I ended up doing informational interviews with friends of friends who were kind of in this industry, and also taking online courses to make sure I understood how to do research in this new setting. I felt I had a very strong foundation in experimental design and doing research, given my PhD. I thought I’d try [on my own] and if it doesn’t work out then maybe I’d do some other path—for instance, a bootcamp dedicated to UX or go back to get some kind of degree that’s actually in UX. But I ended up not needing to take that path.

RH: Do you have advice for grad students who want to move into new areas such as industry or tech?

CL: It was quite a learning process. For me it was multipronged. If I was interested in certain dimensions, I’d try reading in those areas, as well as talking to people who are in those areas. It was trying to figure out—what are they looking for? Then I tried to backtrack to figure out, based off of my experience, what are the pieces they might be interested in? For instance, eye tracking is interesting for UX. So I made sure I included eye tracking, even though it was a very minor part of my thesis. But at least the fact that I can analyze that kind of data is useful. I’ve also been upfront—I will mention that I do eye tracking on animals, not humans, because the expectations will be different. So I try to make it a balance between trying to “sell” my past experiences, as well as balancing the truth of what I’ve actually done, and making that transition very clear for people.

RH: Tell me about your current job.

CL: From Augmedix, I ended up doing contract work for Google focused on UX research. Then there was the opportunity to go full-time, so I totally seized that opportunity. I work on Google Photos as one of the UX researchers in a product team. I’m interested in questions such as: given our current products, how are our users using it? What are their reactions to it, their attitude towards the product? What should we be building next? What kinds of problems do our users or potential users have—what can we do for them?

RH: How do you go about answering these questions?

CL: The methods range quite a bit. It depends on the question. I work really closely with product managers and designers, sometimes the marketing team, sometimes the data analytics team, and then a minor part with the engineers—it depends. If the question is about something that is not working in our product, then it might be very narrow. I might be going to usability studies that don’t require a lot of participants. It might just be showing them a prototype and watching them use it. It could be done remotely if I am interested in our different markets. Most of our users are not in the US, so if I’m interested in somebody in Europe, I might do a remote study, or we might actually visit people.

Some other questions we’ve been looking at are: you take a lot of photos when you’re on trips, what does that experience look like? It’s really hard for us to follow people on their trips, but sometimes we try to approximate it. For one project, we shadowed people in Disneyland in Tokyo and observed—what are their pain points? It’s really hard for people to articulate what their problems are or what their quirks are, because it’s done subconsciously. So we try to make sure that we can see that kind of behavior, in addition to talking to them, to see what they think their problems are. We also do a lot of surveys.

RH: What do you enjoy about your work?

CL: I love my current team. I like making sure that I have an environment where people are interesting and not bored, so there’s a really good, stimulating team and also the subject matter is really, really interesting. I’ve always liked people watching. If you study photography you can constantly do research. In fact, if I sit outside, there’d probably be somebody taking photos or struggling to share photos. It’s been really nice because my research can go pretty much anywhere. One perk about being on my team is that whenever we travel, we can do research wherever we go. Also, we have the opportunity to plan studies abroad where we bring the members of our team. We planned a trip to different countries to just see what our users and what photography looks like in these countries. That’s been really fun.

RH: Do you have any hobbies, things you do for fun outside of your job?

CL: Apparently I really love my work (laughs) so I consult for UX research—I’m helping out a startup. Aside from that, I do a lot of wearable electronics, which was kind of a hobby that started when I was in grad school. I always liked tinkering with things and making things, so I started making a lot of accessories and clothes I would wear at the [Berkeley Neuroscience] annual retreats. I would wear my stuff there and test it out before testing it out in the nightlife scene.

Now I usually just make my holiday dresses because we have an annual holiday party, so I’m always wearing something that I made and it’s usually interactive with some electronics in it. I’ve had dresses that light up in different patterns depending on your heart rate. I also have ones that are responsive to sound so they light up when there’s really loud noise or they might change color—like more red when it is truly loud and more blue when it is quieter. I had a skirt from grad school that would change colors depending on the movement.

There was one dress where you got to vote in a web app for what color it should be. It could be blues or pinks or yellows, and you kept going to the web app to vote. It kind of felt like Disney’s Sleeping Beauty where the two fairy godmothers are trying to decide what color dress Aurora should have (laughs).

Photograph: Luk in her 2018 light-up holiday dress.

RH: Is there anything you would like to add?

CL: If anybody wants to reach out with questions about UX, I am happy to be somebody to bounce ideas off of!

by Rachel Henderson
Original Publication: January 29, 2019

Additional Information