AI leader Lei Yu returns to ASU to help shape the future

Lei Yu still remembers arriving at Arizona State University with little more than a pair of suitcases and determination.

“I only knew two things for sure,” he says. “I was joining a great AI program, and I knew I was going to live in the desert.”

In 2005, Yu earned his doctoral degree from the School of Computing and Augmented Intelligence, part of the Ira A. Fulton Schools of Engineering at ASU, working in the nascent laboratory of emerging AI leader Huan Liu. Then, Liu was a new faculty member building his research group, and Yu was an inquisitive graduate student. Today, Liu is an ASU Regents Professor and globally renowned computer scientist, while Yu is an internationally respected AI industry leader.

This March, after more than two decades, Yu returned to ASU to participate in the “Future of Computer Science” brainstorming session alongside fellow academic and industry leaders.

The timing was significant. As AI rapidly reshapes industries, both education and the workforce are confronting difficult questions: What skills will engineers need in 2030? How should universities prepare students for a future where AI is embedded into nearly every profession? And how do organizations navigate technological disruption responsibly?

In moments of upheaval, steady leadership matters. Yu’s career spans academia, federal research and industry innovation, positioning him to answer pressing questions with rare credibility.

Yu serves as vice president of AI and data at General Dynamics Information Technology, or GDIT, where he leads efforts to scale trusted AI adoption across a workforce of 25,000 employees. Before joining GDIT, he helped transform Expression Networks into an AI-first company, developing agentic systems and securing federal contracts. Earlier in his career, he spent 16 years as a professor of computer science at Binghamton University.

Yu credits much of that steady approach to his mentor, Liu.

Yu was Liu’s first doctoral student at ASU, joining the lab in 2000 soon after finishing his bachelor’s degree in computer engineering at Dalian University of Technology in China.

“I was very lucky on my trajectory,” Yu says. “Dr. Liu set me up for success.”

At the time, Liu himself was just beginning to establish what would become one of the country’s most influential AI research groups. Yu entered the lab as machine learning was still emerging as a mainstream field, focusing on feature selection, a process used to identify which pieces of data are truly important.

Under Liu’s guidance, Yu quickly found success. His first research paper was accepted to the International Conference on Machine Learning, or ICML, one of the field’s premier conferences. The work, which explored how to improve learning performance using the smallest and most meaningful set of data features, evolved into Yu’s dissertation and has since earned more than 10,000 citations.

But Yu says the most lasting lesson from Liu was not technical.

“One key idea Dr. Liu always emphasized was ‘slow is fast,’” Yu says.

Rather than chasing rapid recognition, Liu encouraged Yu to focus on building strong fundamentals, producing distinctive work and developing depth as a researcher.

Liu says that patience became one of Yu’s defining strengths.

“Lei distinguished himself early as someone who valued depth over speed,” Liu says. “He built his expertise carefully, with patience and rigor, and that foundation ultimately allowed him to thrive as both a scholar and an industry leader.”

Yu carried that philosophy throughout his career. After earning tenure at Binghamton, he began thinking more seriously about how AI research could move beyond theory and into large-scale real-world application.

“I loved research and teaching,” Yu says. “But after tenure, I started wanting to see the work applied in ways that could impact people more directly.”

The transition to industry was deliberate rather than abrupt. Yu first joined the smaller company Expression Networks before eventually stepping into executive leadership at GDIT.

There, Yu drives enterprise AI strategy and leads efforts to scale what he calls “AI adoption with a soul,” still guided by the same steady philosophy that began in Liu’s lab more than two decades ago.

Rather than viewing automation as a replacement for workers, Yu argues organizations must prepare for a hybrid workforce where humans and AI systems operate side by side. He believes the transformation will depend on how effectively people embrace collaboration.

That perspective shaped many of Yu’s contributions during the ASU brainstorming session. Yu said that the ideal 2030 graduate is someone who views AI as a teammate — whether as an assistant, advisor or cross-functional peer — and can thrive in an AI-enhanced environment.

He emphasized that technical skills alone are not enough.

“If you look at the things you learn as a tree, tools and programming languages are seasonal,” Yu says. “Then there are the roots: computational thinking, communication skills and interdisciplinary understanding.”

Yu believes communication will become especially important in the AI era because interacting with AI increasingly resembles collaborating with people, including setting goals, refining prompts, evaluating outputs and iterating collaboratively.

“The best communicator with people will be the best with AI,” he says.

He also encourages students to stay intellectually curious and ask difficult questions.

“Back when I was teaching, the best students were the ones who asked interesting questions,” Yu says.

For Ross Maciejewski, director of the School of Computing and Augmented Intelligence, alumni like Yu represent exactly the kind of partnership universities need as technology evolves faster than ever.

“Collaboration between academia, industry and our alumni community is essential,” Maciejewski says. “People like Lei help us understand not only where computer science is going, but how we can prepare students to lead responsibly in that future.”