Nam Hai first learned about AI from his chemistry teacher during his 11th-grade year at High School for Gifted Students, VNU Hanoi, even though AI tools were not as widespread then as they are today.

However, it was not until 12th grade when seriously selecting a major that the young man researched deeper. Realizing that the major leaned heavily toward Mathematics and programming - matching his strengths - Hai chose Data Science and AI at HUST, the most prestigious polytechnic school in Vietnam.

During his early university years, Hai spent significant time on general courses, particularly those related to Mathematics. He early recognized that pursuing AI seriously meant he could not underestimate this foundation.

Compared to high school, the larger volume of knowledge and a learning style demanding high autonomy required some time to adapt. Beyond class hours, Hai self-studied via YouTube, read books, and searched for problem sets online to practice. As a result, in Math and Programming courses, Hai consistently achieved A or A+ grades.

Starting his second year, Hai joined an AI research lab at the university under the guidance of Dr. Ngo Van Linh and Assoc. Prof. Than Quang Khoat. During that period, whenever he had no classes, Hai would spend his time in the research lab.

Initially, his advisor usually assigned Hai to read scientific papers and re-program the algorithms presented in them.

In his third year, Hai was admitted to VinAI Research's AI research program. Because he remained a student, Hai was allowed to work part-time. On days without classes, he went to the company to conduct research.

This environment helped Hai recognize the gap between lecture hall knowledge and practical research. Many terms and concepts in papers remained unfamiliar, while AI evolved rapidly, forcing him to continuously update himself.

About six months after joining the company, Hai submitted his first paper to an international conference. The paper was rejected.

"At that time, I was quite sad and stressed, even wondering whether I was truly suited for a research path," Hai recalled.

Instead of giving up, Hai and his collaborators continued running experiments, revising, and resubmitting. This time, the paper was selected as one of the spotlight papers at that year's conference.

"This experience helped me understand that rejection is a normal part of research. The key is knowing how to control emotions, reflect on missing points to improve, and not letting a single failure ruin motivation," Hai said.

Hai's learning and research approach gradually shifted from "learning to understand what already exists" to "learning to ask new questions."

The nod from American universities

During his undergraduate years, Hai co-authored three research papers published at ICLR and ICML, two of the world's leading, top-tier international conferences on AI and machine learning. In all three papers, Hai served as co-first author.

Even so, in his early university years, Hai had not thought about pursuing a PhD abroad. But the deeper he delved into research, the more he realized that many individuals pursuing deep AI continued to higher education.

In late 2025, Hai began preparing his study-abroad portfolio and subsequently underwent interviews with universities and professors. Most schools conducted one to two interview rounds. At Cornell University, Hai went through three rounds featuring questions ranging from programming, Mathematics, and AI to his past projects.

Around mid-February 2026, Hai received an acceptance email and scholarship offer from Cornell University. Afterward, multiple universities sent acceptance letters in succession. In total, Hai was admitted to PhD programs at nine US universities, including Virginia Tech, Vanderbilt University, Rice University, Northeastern University, New York University, UT Austin, Johns Hopkins University, UC San Diego, and Cornell University.

Among them, Hai received honor scholarships from Cornell, Vanderbilt, and Northeastern due to his outstanding achievements.

The HUST alumnus decided to choose Cornell, which granted him a financial support package with a total value of $432,895 over 5 years (around VND11.5 billion), covering tuition, health insurance, and an annual living stipend.

Currently, Hai is a first-year PhD candidate at Cornell. The research direction the Vietnamese youth is pursuing focuses on gaining a clearer understanding of how AI operates internally and why a model yields a specific answer or decision.

Thuy Nga