Ky-Vinh Mai was born in a Vietnamese refugee camp in the Philippines. When his family relocated to the United States, they moved around in Orange County until finally settling in Santa Ana. As a first-generation child of refugees who worked hard to provide for their family, Mai learned from an early age how to be independent and resourceful.

“It was just me and my family, and that taught me to figure things out on my own,” the Regents’ scholar says.

These lessons of perseverance and self-reliance became the foundation of his academic success and fueled his curiosity about the world.

Now a senior at UC Irvine double-majoring in data science and language science, Mai’s path reflects his commitment to exploring big questions—ones that don’t always fit neatly into a single discipline. From his early interests in artificial intelligence (AI) to his research at the intersection of linguistics and machine learning, Mai’s academic career has been anything but conventional.

Discovering the value of language science

When Mai first arrived at UC Irvine as a freshman, he knew he wanted to pursue research in artificial intelligence. As a high school student, he participated in Achievement Institute for Stem Scholars (AISS), a college prep program where he conducted a research project on AI automation over the summer. This experience left him wanting to do more. At UC Irvine, studying data science seemed like the natural path, but he found himself concerned about questions relating to human impact. 

“A lot of the advice I heard was to focus on math and stats,” he says. “But I felt frustrated because it left out the human and social dimensions of technology.”

That frustration eventually led him to the language science department, where he found a more interdisciplinary approach to AI. Language science combined his technical skills with an understanding of how humans communicate. It also introduced him to the potential of natural language processing, or the field of study that helps machines interact with human language in the way humans naturally communicate.

“I wanted to see how AI could be inspired by human cognition,” he says. “How do people understand language? How do we resolve ambiguity in speech? Those questions were fascinating to me.”

One of Mai’s first projects as a student researcher involved identifying ambiguous sentences in speech. Working with transcripts and audio from NPR podcasts, Mai built tools to extract examples from millions of sentences.

Language science associate professor Gregory Scontras remembers when Mai contacted him early in his time at UC Irvine.

“I’ve known Ky since he joined UCI and reached out to me to talk about language science,” says Scontras, who is also the department’s undergraduate director. “He immediately began participating in department events and lab meetings and quickly got involved in research.”

Building bridges across disciplines

What makes Mai stand out isn’t just his technical ability—it’s his belief that research should tackle real-world challenges. He has a particular interest in the ethical and social implications of AI.

“Technology is moving so fast, but we aren’t always equipped to regulate it or think about how it impacts people,” he says.

Mai’s approach blends his technical expertise with a sense of social responsibility. He points out that biases in AI systems can reinforce inequalities and that researchers need to think critically about diversity and representation in their work. He credits the language science department with helping him develop this perspective.

“Traditional linguistics used to be more theoretical, but now we’re focusing on actual language as it’s used in the world. It’s a new frontier, and it’s exciting to be a part of it,” he says.

Scontras sees Mai’s potential to make a lasting impact in his future career.

“It’s clear that he’s passionate about research, and his trajectory is very obviously leading him to become an independent researcher,” he says.

A question of opportunity

Mai’s willingness to put himself out there has led to some of his most rewarding opportunities. During his sophomore year, he attended a Ph.D. dissertation defense during finals week, where he realized he was the only undergraduate there. The presenter, Arseny Moskvichev, was so impressed by Mai’s questions that he invited him to collaborate on a project. When Moskvichev later became a research fellow at the renowned Santa Fe Institute, he invited Mai to join him as an intern and researcher.

“It was an incredible experience,” he says. “The Santa Fe Institute is all about interdisciplinary research, and I got to work with people from all kinds of fields. It was amazing to see how different perspectives come together.”

Mai's impressive technical background, strong academic record, and experience with advanced AI research landed him an internship at Los Alamos National Laboratory's Open Science division. There, he focused on addressing critical challenges in climate modeling and materials science, working specifically on identifying and mitigating pretraining bias in AI systems.

"I chose to work specifically on climate science and material science, and to address pretraining bias in AI systems—work that could help us better understand and address climate change and the machines that model it,” Mai says. “But I also recognize that we can't entirely separate our work from its institutional context, legacies and mixed funding sources. These are the kinds of difficult questions that scientists and researchers need to grapple with openly."

Looking ahead

As Mai prepares to graduate, he’s already looking toward the next challenge. His long-term goal is to pursue a Ph.D. and become a research scientist, but he’s also eager to explore new fields. Lately, he’s been drawn to robotics and the emerging field of embodied AI, which looks at how machines can understand and interact with the physical world.

“What excites me most is combining my knowledge of language science with robotics,” he says. “There’s so much potential in how humans and machines can communicate. This brings about complex problems, but these are the kinds of challenges I love.”

Mai’s curiosity and drive have already taken him far, but he approaches the future with humility and an open mind. Reflecting on his time at UC Irvine, he shares, “I’ve learned that so much can happen if you just ask. If you’re open to exploring and willing to take risks, you never know what opportunities might come your way.”

With his blend of technical skills, interdisciplinary thinking, and a deep sense of purpose, Mai is poised to make a meaningful impact—whether he’s solving complex AI problems, bridging disciplines, or asking the next big question.

-Jill Kato for UC Irvine School of Social Sciences
-pictured: Ky-Vinh Mai, presenting "Visual Analogy Making with Large Language Models" at the Santa Fe Institute.