Ruth Fong: Teaching the creativity of computer science

By Julia Schwarz

For Ruth Fong, the fundamental appeal of computer science lies in its creativity. “To be able to program,” she said, “is to be able to create and be creative.”

Ruth Fong. Photo by David Kelly Crow

An emphasis on creativity has been a guiding theme in her work — first as an undergraduate at Harvard, then as a Rhodes Scholar and doctoral student at Oxford, and now as one of 12 members of the teaching faculty in Princeton’s Department of Computer Science.

Fong’s interest in programming was sparked in an introductory course in college, where for her final project she built an iOS app that allowed her to order takeout meals online from the dining hall. “I thought it was really neat that I could identify this need and then solve it by building technology,” she said.

Teaching faculty are essential to the success of the computer science department and valued members of the faculty, said Szymon Rusinkiewicz, vice chair of the department. "While the number and diversity of undergraduates taking courses in computer science has grown rapidly over the past few years, the teaching faculty enable us to continue to provide an outstanding educational experience for all students,” said Rusinkiewicz.

One of Fong’s contributions has been to incorporate more project-based learning into the curriculum. After joining the department in 2021, she introduced an open-ended final project to the curriculum in COS 126, the department’s introductory course. COS 126 is one of the largest courses on campus, with an average enrollment of 300 students per semester. About half of all Princeton undergraduates take COS 126, according to Kobi Kaplan, the course administrator. Kaplan said the final project, pioneered by Fong, has been “a great success.”

The project gives students a blank slate to be creative, according to Fong. “I really believe strongly that students should feel ownership over their work and knowledge,” she said. One of the most gratifying experiences in teaching, she said, is when a student learns so much in their final project that they become the teacher. “You see a light bulb go on. To help facilitate someone's learning journey like that is really exciting.”

Fong specifically sought out teaching jobs after finishing her doctoral work at Oxford, and a position at Princeton was particularly attractive, she said, because the teaching load is low enough that teaching faculty have time to pursue other academic interests, a model they call “teaching + x.” In Fong’s case, this means she has time to pursue her research on machine learning fairness — finding ways to mitigate biases in large data sets — and methods for providing accessible explanations of complex AI models. 

Since arriving at Princeton, Fong has collaborated with Olga Russakovsy’s lab and co-authored several papers. One recent paper, a collaboration with Russakovsky, looked at gender bias in machine learning models. Current approaches to mitigating bias focus on confusing AI models to create a gender-blind effect. Fong and her colleagues showed that gender artifacts are so pervasive in visual datasets that a gender label can be predicted using only the average color of an image. Given this, the researchers suggest that data sets should be aware of information like gender, rather than blind to it. “Instead of confusing our model, we should make it robust and aware of protected classes in order to perform well across the data and demographic groups,” Fong said.

Nicole Meister, first author on the paper, said that Fong’s expertise in explainable AI added a depth of knowledge to the paper that would have been missing otherwise. Meister graduated from Princeton with a bachelor’s degree in electrical and computer engineering in 2021 and is now a Ph.D. student at Stanford. She was also “super inspired” by Fong’s passion for teaching. “She has a lot of goals for improving computer science education,” Meister said. “I was really excited when she joined the department.”

In the spring, Fong will teach “Introduction to Machine Learning” (COS 324) for a second time. "We're thrilled,” Rusinkiewicz said, “to offer a course on machine learning taught by someone who is not only an amazing instructor, but also a world expert in explainable AI."

Many students have no exposure to the field before they come to college, Fong said, and starting out can be like learning a new language. We do a lot of this kind of learning early in our lives, said Fong, who also has a background in neuroscience, but it gets more challenging as we get older. “Showing undergraduates that they can still learn something completely new is really rewarding.”

The Department of Computer Science is currently seeking applicants for teaching faculty positions from exceptional individuals who share a strong commitment to undergraduate education.