Inside China’s top school training the next AI engineers

Felix Gunn, a second-year doctoral computer science student, said he essentially lives a “996” routine. Tsinghua University in Beijing. Seven days a week, his day begins around 8 a.m., when he leaves campus. Around 9pm

Most of that time is spent on research, but he’s not just working toward a degree. After graduation, Gan hopes to develop technology that can be turned into a product and launch a startup.

This level of enthusiasm is the norm on campus, and writing one paper is not enough to stand out. Some students write more than one.

Tsinghua University is China’s top computer science university and is often compared to Harvard University and the Massachusetts Institute of Technology. Students are preparing for a front-row seat in the global AI race.

As China’s AI ambitions grow, so does the pressure on training its engineers and researchers.

Many say the payoff for years of hard schooling is worth it. Passing through Tsinghua could mean a golden ticket to the ranks of China’s tech elite. The network includes Wang Xing, co-founder and CEO of Meituan, one of China’s largest on-demand service companies. Other notable alumni of Tsinghua University include Chinese President Xi Jinping and his predecessor Hu Jintao.

According to market intelligence platform Tracxn, Tsinghua University graduates have founded more than 1,000 companies, including more than 10 unicorns.

Normalized “shattering” culture

This university is one of the most selective universities in China. To gain admission, students must achieve top scores on Gaokao, China’s national university entrance exam.

According to national data cited by Chinese media, in 2024, only about 0.05% of applicants to Tsinghua University and Peking University will pass the high school entrance exams, or about 1 in 2,000.

Most of Tsinghua University’s undergraduate students come from mainland China. Graduate schools, such as Tsinghua University’s Shenzhen International Graduate School, have a large number of international students. Located approximately 1,400 miles south of Beijing, the graduate school attracts Chinese and international students in research-driven, industry-linked programs.

Daniyal Kuzekov, a master’s student studying data science and information technology at Tsinghua University’s Shenzhen Graduate School, said most of the students he meets either have strong backgrounds or are relentless in their efforts.

“I feel the pressure they’ve been under and I feel how hard they’ve worked to get to this level,” he added.

Kuzekov, a native of Kazakhstan, talked about working with an AI assistant to write a paper on DeepSeek, and going on a 22-hour coding marathon to write the paper.

Students told Business Insider that learning at the frontier of AI comes with a culture of grind. Computer science undergraduate Pau Tong Lin Xu typically studies seven to eight hours a day until late at night.

“You learn a lot, but it comes at the cost of a huge amount of work, which can leave you with little time to do other things or explore your ideas,” he says.

He said peer pressure is a “double-edged sword”, both motivating and sometimes overwhelming.

Tsinghua’s culture reflects broader trends in China and is often described as “innovative.” This is a cycle of intense competition where everyone works hard to keep up.

Some young Chinese people resist by “laying down” and refusing to do hard labor. But for students at top universities, opting out is rarely an option.

Publish early, join labs, and code as a baseline

Many computer science students take time out of their day-to-day responsibilities to join a lab early on and work on academic papers.

Some undergraduates go a step further and aim to present at prestigious computer science conferences. This is an indicator of your research ability when applying to graduate programs.

Zhixun Tan, who graduated with a bachelor’s degree in electrical engineering in 2016 and works as a software engineer at Google, said participating in labs and participating in activities outside the classroom were the most valuable parts of his Tsinghua experience.

The professors were happy to guide the students’ research programs. “If you just contact a random professor, most of the time they’ll be happy to take you there,” Tan said.

Ms Tan completed a summer research program organized by her school at Imperial College London, which she said broadened her research horizons.

Davenzka Abigail, a sophomore majoring in biomedical engineering, said that at Tsinghua University, coding is treated as a fundamental skill for the entire field. Economics students might start using Python for data analysis. Science and engineering students often learn languages ​​like C++ early on and learn them more deeply, she said.

Abigail completed a C++ course in her first semester.

“It helps in the long run,” she said, adding that coding has helped her with classes and projects.

Students also use AI tools every day to enhance their learning.

For Gunn, Ph.D. AI tripled the speed of his students’ research, helping with pre-checking, code drafting, and early implementation. However, the last step remains manual.

“Professors tell us to use the tools and not to resist them, but we need exact precision in our research, so we have to manually verify the code at the end,” he said.

Kuzekov said he uses AI tools in his machine learning courses with the approval of his professors.

“We need to share chats and all these materials with teachers,” he said.

In some classes, students reveal how much AI is involved in their work. Abigail said professors can also cross-check usage to validate their research.

Zach Dichtwald, founder of the Shanghai-based think tank Young China Group, said Tsinghua University’s reputation comes from being “China’s most selective, best-resourced and most symbolically important university, with an especially strong engineering and computer science pipeline leading directly to elite jobs and AI research.”

“There is a growing recognition that China is the world’s leading source of AI talent, and visibility is even more important than before,” he added.

ticket to the top

For many students, the benefits of attending Tsinghua University become immediately apparent.

Brian Konstantin Sadihin, a second-year master’s student studying computer science, said Tsinghua’s training gave him an edge in competitive technical interviews.

A year into the program, he secured a research internship at ShengShu, one of China’s leading AI startups, where he worked on multimodal generative AI models. He said Tsinghua’s reputation and training made him “optimistic” about his job prospects.

Technology roles are the top destination. Approximately one-third of Tsinghua University’s undergraduate students will be employed in IT and software-related industries in 2022, according to the university’s data.

Qiao Litao, a senior researcher at the Institute of East Asian Studies at the National University of Singapore, said Tsinghua University is at the pinnacle of higher education in China.

“Its reputation reflects not only its academic strength but also its historical role in cultivating political, scientific and technological elites,” said Zhao, who studies Chinese social policy.

Mr. Zhao said Tsinghua University graduates typically work for “major technology companies or high-growth startups” or pursue further studies. Recommended technology companies include Huawei, ByteDance, Tencent, and Alibaba.

With geopolitical tensions rising and local opportunities expanding, academia remains an important avenue, and many will remain in China, Zhao said.

“Notably, a small but influential segment is moving towards AI-focused startups, often working with university labs,” he added.

A front row seat to realize China’s AI ambitions

As China closes the gap with the U.S. in AI and challenges some areas, Tsinghua University students find themselves at the center of AI.

Tan, a software engineer at Google, said that even as a college student in the mid-2010s, it was clear to me that China was at the forefront of technology and was rapidly producing technology talent.

NUS researcher Zhao said China’s university system is increasingly structured to direct talent to strategically important areas, and AI is one of the top priorities.

Fueled by strong state support and rapid industry growth, the sector is widely seen as a frontier of both scientific progress and economic opportunity.

“This is not a purely top-down process; student preferences, labor market signals and organizational incentives are mutually reinforcing, especially in AI,” he added.

For some Tsinghua University students and alumni, being part of China’s top AI training ground comes with a sense of responsibility to lead the technology revolution.

“No one can solve the problem except us, so we have to step forward and tackle the hardest things, so others will follow,” Tan said.