China's STEM education makes Americans feel anxious

The source of anxiety: the visible gap is widening

American anxiety is not groundless.

The data-driven comparisons are striking: by 2026, China’s annual college graduate population is projected to reach 12.7 million; STEM (science, technology, engineering, and mathematics) programs alone produce over 5 million graduates each year, with roughly 1.3 million bachelor’s degrees awarded in engineering annually. According to data from the National Center for Education Statistics, the number in the United States is about 130000 to 145000-a gap of nearly ten times, making the "engineer dividend" no longer just a slogan.

At the same time, basic education in the United States continues to decline.

Some California schools have put off teaching algebra because students are not ready, and even Harvard University has begun offering remedial math courses for freshmen. PISA data show that China's four provinces and cities rank first in science, mathematics and reading, while the United States ranks 13th in science, 18th in mathematics and 37th in reading. Even the US congressman said bluntly: "all Chinese students have higher scores in mathematics and natural sciences than American students".

Chinese science education, really "crush" the United States?

The truth may not be so simple!

The Global STEM Education Development Index shows that the United States still ranks first with 86.50 points, followed by China with 85.46 points. The advantages of the United States are in education investment, resource allocation and the agglomeration effect of top universities, while China's advantages are in teachers' ability, competition results and graduate scale-one is "resource driven" and the other is "efficiency driven".

WAIC A set of data revealed on the 2026 also confirms the transformation of this "efficiency advantage": the cumulative download volume of China's AI open source model has exceeded 10 billion times, ranking first in the world; foreign media analysis pointed out that Kimi K3 is the representative,The gap between the US and China's frontier AI models may have narrowed to "two or three months".

But at the same time, we must also see that the investment in the 32 key projects signed at this conference exceeds 40.9 billion yuan-behind the huge investment is the reality that we still need to catch up with in original innovation and basic theory.

The Fields Medal and Kimi K3 are to be proud of, but they represent a breakthrough in the "point" rather than a qualitative change in the "face. From "follower" to "leader", there is still a long way to go.

Two educational logics, each with its own anxiety

The strengths of science education in China are familiar to everyone--The knowledge system is tight, the training intensity is high, and the foundation is solid.From elementary school mathematics to college entrance examination, a set of mature selection and training mechanism ensures that a large number of students reach a higher level of mathematics. At the family level, many American parents may be surprised by the importance Chinese parents attach to their studies; hard work, self-discipline and persistence are widely regarded as essential qualities for success.

But the coin has another side, which is also the deep anxiety of Chinese educators:This system is good at cultivating "problem solvers", but not necessarily good at cultivating "questioners".Our students can do almost anything in the exam, but outside of the classroom, how many of us really love science, dare to question authority, and enjoy the process of exploring the unknown? This anxiety is further amplified when AI can solve the math problems of the college entrance examination in seconds-how much of the irreplaceability of our proud "problem-solving ability" is left?

America's anxiety comes from the other direction. Anduril founder Laki was outspoken in an interview with Fortune: American colleges and universities are increasingly divorced from engineering practice, teaching students to "think big" but forgetting to teach them "down to earth". He cited Apple's changes in recent years as an example-in the past, Apple needed to solve its own engineering problems in manufacturing products, and now "most of the most difficult work is done by Chinese engineers". The anxiety of Americans is that they have no innovative ideas, but they have lost the engineering ability to land them.

On the one hand, it is solid but worried about "failure", and on the other hand, it is flexible but anxious "hollow". The two anxieties stem from the same era proposition: when AI begins to replace "problem solving" and when global competition shifts from papers to products, what kind of people should our education cultivate?

The technical signal released on the WAIC 2026 may have pressed the accelerator button for both anxieties. The core trend conveyed by the conference is that AI is evolving from "being able to speak" to "being able to do", which also means that future science and engineering talents must not only have a solid mathematical foundation, but also have interdisciplinary integration capabilities and hands-on practice. Ability-and this is precisely the "no man's land" where both education systems feel anxious ".

How can international schools become the "greatest common divisor" of science education between China and the United States "?

Back to the question that international education practitioners are most concerned about: what does this discussion on science education in China and the United States have to do with us?

There is, and not small.

The trend of WAIC 2026 is already clear-AI agents, artificial intelligence, humanoid robots… The core of these technologies is no longer "how strong the algorithm is", but "whether it can solve real-world problems". This means that the future top science and engineering talents need to have both depth and breadth, theory and practice, logic and creativity. The interdisciplinary curriculum and project-based learning experience of international schools are precisely in line with the training needs of this kind of "T-type talents.

In the past, "climbing the vine" was the mainstream direction of students in international schools, but with the improvement of the quality of Chinese science and engineering personnel training, the enhancement of the international reputation of domestic universities, and the uncertainty of the policy of studying in the United States, more students will look to the science and engineering projects of top domestic universities in the future. The guidance system of international schools needs to start to take this "new road" seriously ".

AP calculus, IB advanced physics, A- Level mathematics… The international curriculum system has advantages in cultivating inquiry ability and global vision, but in terms of the "solidity" of mathematical training, can it satisfy those students who want to go further in the direction of science and engineering? Some international schools have begun to introduce the methodology of Chinese mathematics teaching, taking the advantages of the two. The enlightenment of WAIC 2026 is: if future AI engineers need to "really work" in the physical world, then our curriculum can't just stay on paper-programming courses, do we need more experimental courses, project courses, and interdisciplinary workshops? This kind of exploration deserves more attention from schools.

At the WAIC 2026, 29 countries jointly signed an agreement to establish the World Artificial Intelligence Cooperation Organization; China announced that it will provide 5000 AI training places for developing countries in the next five years. AI governance, AI ethics, and the global fair distribution of AI… these issues are no longer the "icing on the cake", but the core propositions of the AI industry. People who can understand and participate in global conversations will be scarcer than pure technical talent.

If Chinese science education has advantages in "depth" and American science education has advantages in "breadth", then the most important thing for international schools is not to choose between the two, but to be the place that can provide two kinds of nutrition at the same time. This is not a "shortcut to escape the college entrance examination", but an educational experiment to explore whether it is possible to cultivate people who have a solid mathematical foundation, interdisciplinary vision and original thinking.

In addition to anxiety, someone is trying different ideas to solve the problem.

Whether Americans are anxious or not is not that important. What is really worth asking is: what kind of people do we want to cultivate in AI era from "being able to speak" to "being able to do?

There are already international schools trying to answer this question.CRESTAwardsEstablished in 1986 by the British Science Association, the global STEM education accreditation system offers students aged 5 to 19 a complete inquiry-based learning path from primary school to Senior high school. Its core idea is simple: let students think and act like scientists-choose topics independently, design experiments, analyze data, and do real research instead of brushing questions.

This just responds to the two anxieties mentioned above: to retain the "solid" of Chinese education and to supplement the ability of "asking questions" and "landing. CREST has been officially recognized by UCAS, and top universities such as Oxford and Cambridge use it as an important reference for STEM professional selection.

The Fields Medal, the AI Breakthrough, the WAIC-these are the transcripts of science education in China, but they are far from the end. What international schools can do is to find a way of integration between the two kinds of education. CREST, perhaps, is one of the pedals worth stepping on.

It's not an easy road, but it's worth it.