Mei Yonghong on Scaling Discovery

BGI executive & former Ministry of Science and Technology official calls for big platforms, big datasets, and bigger teams in a private-sector lead and a national wager on bio-AI

The 20th Central Committee of the Communist Party of China (CPC) convened its fourth plenary session in Beijing from October 20 to 23 and approved its recommendations for the 15th Five-Year Plan; formal passage of the plan must wait til the National People’s Congress in March. In the meantime, we are offering some interesting expectations published on 10 October by BGI’s Mei Yonghong.

Mei Yonghong is Director and Executive Vice President of China’s private BGI Group, one of the world’s leading life science and genomics organisations that, in recent years, has become a prime target in the U.S. The Center for Security and Emerging Technology, a Georgetown University think tank, described BGI as “taking on the same role in the biotech space as Huawei has in telecommunications” in a 111-page report in May 2024.

Immediately before joining BGI, Mei was Mayor of Jining, Shandong Province, between 2010 and 2015. Before that, he served as Deputy Director-General of the General Office, Director-General of the Research Office, and Director-General of the Policy, Regulation, and System Reform Department of the Ministry of Science and Technology (MOST).

China’s most impressive advances, argues Mr Mei, have come less from faculties than from factories: a pipeline of scientists, a vast manufacturing base, and an ocean of data that turn half-formed ideas into products at scale. The next chapter should entrench that practice. Organise science along engineering lines and put laboratories in the production flow. Let businesses, NOT universities, be the principal innovators. Build big platforms, big datasets, and big teams. Treat AI as a co-scientist and data as a strategic asset.

Mei also argues that the fusion of AI and biotechnology is the next frontier. The coming industrial revolution, he says, will shift the locus of progress from inert matter to organic life and the human genetic code. “When disease can be detected and treated early, when food can be synthetically produced, when energy and materials can be ‘grown’, and when life can be purposefully modified or even remade, the future will diverge sharply from the past. No historical transformation is comparable; the world now stands at the leading edge of an unprecedented change.”

—Yuxuan Jia

The following remarks were originally published on the official WeChat blog of Huagu Biotechnology and Bioindustry Research Institute, a non-governmental, non-profit thinktank set up in 2023 by biotech and agricultural companies in China, including BGI’s affiliates.

梅永红：“十五五”规划的科技五大新背景

Mei Yonghong: Five New Background Factors for sci-tech in the 15th Five-Year Plan

The 15th Five-Year Plan for Economic and Social Development is imminent. From every perspective, scientific and technological progress and innovation will be the top priority. Compared with previous plans, the overall context for China’s sci-tech development has undergone profound changes for the next five years and beyond. This new context should anchor China’s forthcoming sci-tech planning.

I. New Dynamics in China-U.S. Sci-Tech Decoupling

In recent years, competition between China and the United States has intensified. While the most visible flashpoints are in trade, supply chains, and talent, the core rivalry lies in sci-tech. I would go so far as to say that the race for sci-tech leadership will be the contest of this century for the two countries and will shape the national trajectories of both.

Over the past two decades, China’s sci-tech has advanced by leaps and bounds, achieving breakthroughs in almost all key fields, and a host of advanced equipment has emerged. Achievements such as the BeiDou navigation system, the Tiangong-1 space station, the lunar exploration program, the C919 large passenger airliner, drones, robots, artificial intelligence, general-purpose software, the semiconductor industry, high-speed rail, new-energy vehicles, third-generation nuclear power plants, a 10,000-meter manned submersible, 5G, supercomputing, quantum communications, tunnel-boring machines, MRI, as well as—on the military side—aircraft carriers, large destroyers, fifth-generation fighter jets, military transport aircraft, early-warning aircraft, new-type nuclear submarines, and the recently unveiled DF-61, JL-1, and JL-3, collectively reflect the enormous leap in China’s science, technology, and industry.

Many indicators also reflect the rapid growth of China’s sci-tech. In 2024, national R&D investment reached 3.6 trillion yuan, second in the world; the full-time equivalent R&D personnel totalled 7.24 million per year in 2023, first in the world; there were 3.351 million valid invention patents and 1.828 million applications, first in the world; paper output reached 1.216 million, first in the world; China accounted for 48.4% of the world’s hot papers, first in the world; it is home to 26 of the world’s top 100 innovative sci-tech clusters, first in the world; high-tech industry operating revenue reached 22.34 trillion yuan, and more than 16,000 incubators have been cultivated in total.

As natural geopolitical rivals, this is arguably the outcome Washington is least willing to accept. Since China joined the WTO in 2001, rather than giving China a ticket to integrate into the global industrial system, the United States hoped, much as it did with Latin American economies, to lock China’s economy and industry into the low end over the long term. As the well-known American columnist Thomas L. Friedman said, “When China sold us only shallow things, politically speaking, we didn’t care whether China was authoritarian, communist, libertarian or vegetarian. It didn’t matter because you were just selling us shallow goods.”

Therefore, starting from Donald Trump’s first term, a broad bipartisan and wider societal consensus has taken shape around using sci-tech to constrain China, and even to pursue sci-tech “de-Sinicisation.” Measures aligned with this aim have often passed Congress with near-unanimous majorities—rare in U.S. politics—including crafting trade and economic rules that limit the development of Chinese industries, erecting sci-tech barriers to preserve a competitive edge over China, and building international coalitions that narrow Chinese companies’ operating space overseas. To date, more than 1,000 Chinese companies and institutions have been added to the U.S. Entity List, encompassing many of China’s leading firms in frontier fields.

Judging from the sanctions imposed on leading Chinese companies such as Huawei and BGI, the United States has mobilised a whole-of-nation effort:

1. Whole-of-government: Beyond the Department of Commerce, the Department of Defence, the Department of State, the Department of Energy, the Department of the Treasury, Customs and Border Protection, and Congress have all become involved;

2. Whole-of-society: Many universities, companies, sci-tech institutions, and non-governmental organisations have echoed the U.S. government’s actions;

3. All dimensions: In addition to trade measures, coordinated actions span investment, finance, education, data, and media;

4. All allies: By pressuring Western allies and other countries and institutions that cooperate with Chinese companies, the United States is mounting a concerted effort to exclude Chinese firms from the field.

All factors considered, the larger trend of China-U.S. sci-tech decoupling is hard to reverse, and the “window period” for external cooperation seen during China’s National Medium- and Long-term Science and Technology Development Plan two decades ago is unlikely to reappear. After decades of tracking, imitation, and path dependence, China’s sci-tech has entered a blind spot, a “no-man’s land,” with fewer peers to track or benchmarks to follow. Even so, confidence has not wavered. On one hand, no individual or country can fully control the world—a basic, widely shared premise in human civilisation today. On the other hand, China is no longer the “sick man” humiliated a century ago; its sci-tech base, industrial system, and talent pool now stand at the global frontier. A clear paradox is: the tighter the chokehold in a given field, the faster China’s capabilities often advance. In the end, the ambition to strangle China through science and technology will be remembered as a historical joke.

II. New Requirements for Adjustments in the Economic Structure

In recent years, China’s economy has slowed, with growth in the three traditional engines—investment, consumption, and trade—declining across the board. This is not a cyclical dip but the accumulation of long-standing structural imbalances arising from the convergence of multiple structural contradictions. Weak consumption, an overinflated property sector, rising local government debt, the stock market’s “3,000-point curse,” fragile private-sector confidence, and widening regional disparities are each problems of formidable scale. Together, they amount to systemic pressures without precedent in modern world economic history—and a grave challenge the 15th Five-Year Plan must confront and resolve.

It must be acknowledged that China’s rapid growth in recent decades has largely followed a short-cycle, high-speed, expedient model shaped by the imperatives of a shortage economy, with investment expansion, scale effects, and a sizeable demographic dividend as the main drivers. Coupled with fiscal decentralisation—under which each jurisdiction finances its own budget—intense competition among more than 30 provinces, over 300 prefecture-level cities, and roughly 3,000 counties has produced a “fiefdom economy,” turning straightforward capacity build-outs into a century-scale surge unseen since the Industrial Revolution. Today, with China’s industrial output accounting for about 30 per cent of the global total and deglobalisation now a standard geopolitical tool of hegemonic powers, this growth model has clearly become unsustainable.

Total factor productivity can quantitatively and objectively reflect the current state of an economy’s structure. Total factor productivity refers to output growth achieved through technological progress and organisational innovation with all factor inputs held constant. It reflects improvements in technological level and resource allocation efficiency. According to statistics, the current growth rate of China’s total factor productivity is only 40 per cent to 60 per cent of that of advanced economies. In other words, with the same factor inputs, China’s output is only about half that of advanced economies. This represents both room for improvement and a potential path into the “middle income trap” that some countries have failed to escape for a century. I have conducted in-depth field studies in some Latin American and Southeast Asian countries. Their per capita incomes were once far higher than China’s, but because total factor productivity rose slowly, they stagnated for a long time after a brief period of growth.

Experience shows that gaps in total factor productivity cannot be narrowed by expanding general investment. They can only come from sustained progress and innovation in sci-tech and from a comprehensive rise in new, quality productive forces. As early as 30 years ago, China formulated the strategy of invigorating the country through science and education, and 20 years ago set the goals of independent innovation and building an innovative nation. This reflects an understanding of the laws of development and a sense of historical responsibility. If tracking, imitation, and investment-led expansion were once shortcuts to rapid development, a major country cannot stop there. Our era still offers vast horizons; it demands new thinking, new paths, and new methods to reach the world’s leading edge. The “no-man’s-land” should be seen not only as a test, but as an opening for transformation and renewal.

For example, Elon Musk believes human civilisation should not stop at Earth and that Mars could be a new home for the continuation and development of humanity, which led to the creation of SpaceX and, in pursuit of that goal, reusable rockets, Starlink, and Starship. Another example is the ocean, which covers 71 per cent of the Earth’s surface, yet 95 per cent of the deep ocean remains unexplored. The physical, chemical, biological, geological, and energy systems found there extend far beyond mankind’s current body of knowledge. Likewise, the phenomena and governing laws of life—including human life—constitute, in effect, another universe: a macroscopic world that outstrips present imagination. Existing theories of the origin of species, neuroscience, emotion, and thought touch only the tip of the iceberg. Looking to the stars is not merely a scientific pursuit; it is a foundation for the future and will help shape China’s international position in the new century.

It bears emphasis that structural problems cannot be resolved through short-term economic measures alone; they depend chiefly on long-term strategy and planning. In particular, sci-tech progress and innovation cannot be summoned simply by ramping up investment. They require years of sustained, cumulative effort before breakthroughs emerge. This, in turn, calls not only for a change in mindset but also for shifts in ways of thinking and value orientation. Ultimately, innovation-driven development rests on an inevitable and far-reaching transformation of institutions.

I was once asked how long such a transition would take. A reasonable estimate is about ten years, which is the typical horizon for many emerging industries to complete a generational upgrade of technologies and products and to achieve broad market acceptance. For China, the greatest source of confidence lies in the breadth and autonomy of its policy toolkit. Unlike countries such as Japan, China exercises full political and economic sovereignty; on matters of national economic security and strategic interest, it does not need to act at others’ discretion. However strong the resistance, the path of economic transformation and the building of an innovative nation will continue unswervingly.

III. A New Paradigm Led By Big Science

The philosophy and methods of modern scientific research have evolved over centuries. Before the seventeenth century, science was largely centred on independent inquiry by individuals or schools of thought—what might be called “small science.” In the eighteenth century, loose learned societies emerged, initiating more institutionalised exchange and collaboration. In the nineteenth century, scientists moved beyond the ivory tower as industrial application became a primary aim of scientific progress. By the twentieth century, science had scaled to national and even international enterprise, with cross-field integration, a stronger technological orientation, and an engineering turn in science becoming mainstream.

This is not a divergence of methodologies but a deepening grasp of natural laws accompanied by rising capabilities of mankind. The universe and life form integrated systems; reductionism—the notion that the whole is merely the sum of its parts—cannot adequately explain complex natural phenomena. As inquiry pushes toward both macro and micro frontiers, the curiosity-driven, decentralised “small-science” paradigm is giving way to a “big-science” model defined by interdisciplinary integration and the organisation of science along engineering lines.

As early as 1999, Professor Tsung-Dao Lee observed: “It is mistaken to think that understanding elementary particles means understanding the vacuum. In such a simplified picture, there would be no dark matter and no quasars. I think the genome is similar: identifying genes one by one does not solve the mystery of life; life is macroscopic. Twentieth-century civilisation was centred on the microscopic, and I believe in the twenty-first century, the micro and the macro will be joined as one. This is not only true for physics; it may also influence biology and the development of other technologies.”

In 2015, Song Jian, former director of China’s State Scientific and Technological Commission, published an article titled “Reductionism and Holism” in Frontier Science. While fully affirming the foundational role of reductionism in industrialization and post-industrialization, he highlighted its limitations, including neglect of information, which stands alongside matter and energy as one of the three basic elements of the universe; neglect of the hierarchical organization of systems, making it difficult to explain why the whole is greater than the sum of its parts; and the assumption of reversibility in natural processes, whereas biological evolution is irreversible and time does not run backward.

Take BGI as an example. This private high-tech company, without star academics or flashy titles and staffed by a group of researchers around thirty, has published 183 papers in the flagship journals Cell, Nature and Science (CNS), surpassing most specialised institutes, and ranks fourth globally among corporates in the Nature Index. This achievement stems from a new organisational model oriented toward large, audacious goals, underpinned by big data, and coordinated through broad collaboration. Without big data, there is no life science; this has been BGI’s central principle along the way.

This model breaks with linear notions of scientific and technological progress, and with the classic principal-investigator (PI) system. It embodies industry–academia–research integration, unites scientific inquiry with technological development and engineering practice, and represents a new paradigm powered by large platforms and big data. Young scholars on this platform work within clearly targeted research cycles, grounded in massive datasets, embedded in highly open international collaboration, and supported by an academic ecosystem BGI has built over more than two decades. While it may lack “star” academics in the conventional sense, a cohort of energetic, imaginative young researchers equipped with leading tools, the largest datasets, and state-of-the-art platforms can, through long-term focused effort, deliver major results.

On this topic, it is natural to recall Qian Xuesen, a man ahead of his time. After returning to China in the 1950s, he made foundational contributions to the country’s first atomic bomb, intercontinental ballistic missiles, and satellite programmes. He also devoted his career to the study and application of engineering cybernetics, aiming to clarify the internal relationships among science, technology, and engineering and advocating for the technical sciences to lead disciplinary development and technological progress. A contemporary parallel is Elon Musk: his disruptive innovations cut across traditional disciplinary boundaries, overturn the linear “science-technology-product” model, and rebuild R&D logic from first principles.

This pattern reflects deeper laws and trends. Even before the Second World War, the United States had begun to form a Big Science paradigm. The Manhattan Project, the Apollo lunar programme, the Information Superhighway, and the Human Genome Project exemplify large-scale scientific enterprises that crossed disciplinary boundaries and moved beyond linear models of progress. Crucially, they brought scholarship out of the scholastic cloister and aligned it with economic and social needs. The fact that humanity generated more knowledge in the last century than in the previous two millennia owes much to this paradigm.

Many have read Vannevar Bush’s Science, the Endless Frontier, a report that emphasised the importance of basic research and free exploration and had a major impact on postwar U.S. sci-tech policy. But to infer from this that basic research is the starting point of scientific progress or the principal cause of America’s technological preeminence today does not square with the historical record. The United States already led the world in technology before the Second World War, while the global academic centre remained in Europe, which produced roughly 70 per cent of Nobel laureates. Even today, despite hosting the largest number of top universities, the United States continues to fund mission-oriented national laboratories and to support enterprise-led research and innovation. At a minimum, this shows that neither basic research nor technological R&D exists in isolation: basic research depends on technologies, instruments, and data, and scientific advance is embedded in technological development and engineering practice.

It is particularly noteworthy that the rapid rise of AI and large-scale models is overturning the traditional research paradigm, even rendering some research activities obsolete. In protein structure prediction, for example, more than fifty years of global effort have fallen short of even one-thousandth of the output produced by AI models in a single year. As AI diffuses, decentralisation is becoming the norm; research will no longer be the exclusive preserve of high-brow elites; and the age of the lone scientific hero is quietly giving way to specialisation and large-scale collaboration. The result is a growing cohort of “blue-collar” scientists who are likely to become a principal force in discovering new knowledge and creating new industries.

Under the big-science paradigm, research outcomes will no longer be confined to what is often called “from zero to one.” The process “from one to one hundred” is equally a path of discovery and accumulation. Human knowledge comprises explicit and tacit forms. If “zero to one” denotes explicit knowledge, then “one to one hundred” corresponds to tacit, unstructured knowledge. Like an iceberg, explicit knowledge is the visible tip above the waterline, while tacit knowledge is the far larger mass below. Acquiring tacit knowledge is a process of cross-disciplinary integration and engineering practice; it can be achieved only through a Big Science paradigm in which science, technology, and engineering are tightly integrated.

The Big Science paradigm brings not only changes in research methods, but also a profound transformation of research organisation, the science-and-technology system, resource-allocation priorities, and scientific culture. It amounts to a wholesale restructuring of the research enterprise. For China’s sci-tech, this is less a choice than an imperative to embrace.

IV. A New Stage for Enterprises as the Main Innovators

China’s recent Victory Day parade left many proud and encouraged by clear advances in military equipment. The successive induction of next-generation systems, including fifth-generation fighters, new intercontinental missiles, laser weapons, unmanned underwater vehicles, unmanned aircraft, and information systems, signals that the PLA’s equipment level has moved to the global forefront. In contrast to the passive and awkward situation during the Taiwan Strait crisis three decades ago, China today is fully capable of safeguarding national security and dignity.

Why has China achieved such notable results in a field tightly constrained by the West? I share the view that the source lies in strong engineering capability—the capacity to take disparate, cutting-edge and even not-yet-mature technologies and, through system design, integration, testing and verification, and iterative refinement, deliver outcomes within defined cost and time. While gaps remain vis-à-vis Europe and the United States at the scientific knowledge frontier, the application-side engineering capability China has built and the extensive body of know-how accumulated as a result are unique in the world today. This is combinational innovation across disciplines and industries, rooted in a robust industrial ecosystem and oriented toward solving real-world problems. It is the defining strength of Chinese-style innovation.

Behind this capability lie more than seventy years of accumulated technology and talent, the coordinated strength of a fully integrated industrial chain, and the confidence that comes with being the world’s largest industrial nation. For example, more than sixty million business entities provide production, manufacturing, and service capacity across virtually all fields and categories; over twenty million highly qualified engineers drive continuous iteration and upgrading of technologies and products on the production frontline; more than one billion digital citizens generate new demand and endless positive feedback for front-end technological innovation; and the world’s largest, most complete manufacturing system equals the combined capacity of the United States, Japan, Germany, and South Korea.

In fact, over the past two decades, cluster-based innovation and application-led breakthroughs in China have emerged not just in defence sci-tech but also in new-energy vehicles, 5G, artificial intelligence, unmanned aerial vehicles, biomedicine, and genetic engineering. This is a body of knowledge forged by practical needs on the engineering and application side. It forms a vital part of the overall knowledge system and is increasingly becoming a major source of new knowledge and a key lever of economic and social progress. As Friedrich Engels observed more than a century ago, “If society has a technical need, that helps science forward more than ten universities.”

The principal carriers of application-side innovation are China’s rapidly expanding high-tech enterprises. This shift is the most significant change in the country’s sci-tech landscape over the past two decades and the signal achievement of its sci-tech system reforms. In 2024, enterprises accounted for 78 per cent of total R&D spending; among the top ten recipients of invention patents, enterprises made up 70 per cent. Nationwide, there are more than 490,000 high-tech firms; 141,000 innovative SMEs that use specialised, sophisticated technologies to produce novel or unique products, including 14,600 “little giant” firms; 1,557 single-champion manufacturers; and over 570 industrial enterprises in the global top 2,500 companies in R&D investment. With this deeply grounded “wolf pack” model of technological innovation, no technological barrier is insurmountable.

Huawei is the most emblematic case. In 2024, it invested RMB 221.8 billion in R&D, which is 25% more than the Chinese Academy of Sciences and equivalent to the combined research funding of the 76 universities directly under the Ministry of Education. Over more than three decades, the company has stayed tightly focused on mobile communications and built deep capabilities. Under maximum pressure from the United States, a 110,000-strong R&D workforce sustained round-the-clock efforts to clear one technological hurdle after another. In areas such as 5G, semiconductors, operating systems, EDA/design software, and intelligent driving—technological frontiers that the academic community long struggled to traverse—Huawei has delivered breakthrough after breakthrough.

The same holds for new-energy vehicles. A cohort of rising automakers—Chery, Geely, BYD, among others—has fundamentally overturned the long-standing dominance of foreign technologies and brands in China’s auto market. If Western industrialisation was “industrialisation on wheels,” China’s auto industry is now assuming that historic role, too. As Ford Motor Company CEO Jim Farley has observed, 70 per cent of the world’s EVs are produced in China, whose EV technology is well ahead of the United States, with costs and quality far superior to what he has seen in the West. This is disruptive innovation in the fullest sense, surpassing the traditional auto industry from technology to systems, and from infrastructure to consumer models.

Unlike the traditional, project-based academic model, enterprise R&D has two defining features. First, continuity: many firms stay focused on a single domain for ten, twenty, or more years, compounding knowledge and capabilities far beyond what disparate, time-bounded academic activities typically produce. Second, market orientation: science, technology, and engineering are tightly integrated into a closed-loop value-creation system. If scholastic research addresses specific, segmented academic questions, enterprise R&D is organised around competitiveness and delivers definitive outcomes.

Two decades ago, during deliberations on the National Medium- and Long-Term Science and Technology Development Plan, enterprise-led technological innovation was fiercely debated. Many resisted the view that firms should be the principal innovators, treating universities and research institutes as co-actors. Functional convergence across industry, academia, and research, combined with resource allocation that favoured the academic side, left the institutional disconnect between science and the economy unresolved for a long time. Fortunately, as the market economy deepened, enterprises used tangible innovation outcomes to reshape China’s sci-tech landscape.

Recently, Nature published an article about a shift of scientists from academia into industry. Industry-based researchers report markedly higher job satisfaction than their academic peers, and clearer expectations around technological application tend to make results more reliable while allowing researchers to remain patient, confident, and curious. Fei-Fei Li, the “godmother of AI” in Silicon Valley, has likewise noted that the AI research landscape has changed: academia no longer controls the key inputs, including chips, compute, and data, and many of these constraints are resolved more rapidly in the market and in industry.

Taken together, these developments show that enterprises are no longer merely recipients and adapters of technological outputs; they have become organisers of research, leaders in the allocation of sci-tech resources, and the principal innovators. The key swing factor in the Fifteenth Five-Year Plan in Science and Technology will be the extent to which it prioritises and incentivises enterprise-led innovation.

V. New Transformations In the Era Of AI+BT

Recently, the top scientific journal Cell published an article titled “AI mirrors experimental science to uncover a mechanism of gene transfer crucial to bacterial evolution.” Using a large language model (LLM)-based platform, the AI co-scientist, the researchers tackled a problem that had previously required a decade of experimental work. The study shows that AI is not merely an auxiliary tool but a creative engine that accelerates discovery and reshapes how hypotheses are generated and tested. In this case, AI has not only become a genuine “thinking partner,” but has also reached the core experimental finding in just two days!

In this light, the unexpected award of the 2024 Nobel Prize in Chemistry to two AI engineers marks only the beginning of a new scientific era. The Shanghai Artificial Intelligence Laboratory has issued a major review, “From AI for Science to Agentic Science: A Survey on Autonomous Scientific Discovery,” which systematically traces how AI is evolving from a mere tool into an intelligent agent of scientific discovery. The authors argue that AI will not only operate within existing scientific frameworks but also forge new paradigms, invent experimental methods, and even construct theoretical systems. It is poised to act as an intelligent entity that challenges established rules, uncovers deep cross-disciplinary connections, and opens entirely new fields.

Although many uncertainties remain, there is a broad global consensus that AI is poised to rewrite the trajectories of science, economic development, and geopolitics, and the major powers have launched a comprehensive race around it. In the opening to “Winning the Race: America’s AI Action Plan,” the United States declares: “Today, a new frontier of scientific discovery lies before us, defined by transformative technologies such as artificial intelligence… Breakthroughs in these fields have the potential to reshape the global balance of power, spark entirely new industries, and revolutionise the way we live and work. As our global competitors race to exploit these technologies, it is a national security imperative for the United States to achieve and maintain unquestioned and unchallenged global technological dominance. To secure our future, we must harness the full power of American innovation.” It is no exaggeration to say that competition in AI has moved beyond technology into the realm of nations’ fates.

In setting its “AI+” priorities, the United States has placed particular emphasis on AI+BT. The life-science provisions of the recently released “America’s AI Action Plan” should be read as an important bellwether: building world-class biotechnology datasets and explicitly designating high-quality data as a national strategic asset; investing in AI-driven automated laboratories to transform research workflows; and developing more professional and efficient domain-specific models tailored to specific biomedical problems such as protein design, single-cell analysis, drug discovery, and synthetic biology.

In my view, this is not a routine choice of technical pathway but a judgment about where emerging technologies and new relations of production are headed. If, after three industrial revolutions, AI heralds a fourth, its core will point to an interdisciplinary domain: synthetic biomanufacturing. U.S. estimates suggest that manufacturing built on synthetic biology, spanning biomedicine, bio-agriculture, bioenergy, and bio-based materials, could reach an economic scale of around 30 trillion U.S. dollars, effectively the size of another United States.

This is beyond science and beyond the economy. More importantly, the next industrial revolution will move beyond “matter” to focus on organic life and the human genetic code. When disease can be detected and treated early, when food can be synthetically produced, when energy and materials can be “grown”, and when life can be purposefully modified or even remade, the future will diverge sharply from the past. No historical transformation is comparable; the world now stands at the leading edge of an unprecedented change.

This transition will hinge, to a great extent, on the progress of AI+BT. The two are not merely mutually enabling; together, they are redefining the underlying logic. NVIDIA’s Jensen Huang describes this shift as “digital biology,” forecasting it as the greatest transformation ahead. Zheng Hairong of the Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, also argues that artificial intelligence is moving toward biological intelligence. Because at the bottom of it all, the AI era renders data a direct productive force, and the richest, most adaptive data reside within living organisms—organs will become biological supercomputers.

China is not absent from this imminent transformation; it already has a strong foundation and distinctive advantages. In talent, there were about 52,000 AI researchers in 2024, and a pipeline exceeding 700,000, and roughly half of global AI talent comes from China. In research, China accounted for about 60 per cent of highly cited synthetic-biology papers in 2023. In biopharmaceuticals, the number of original drugs from China overtook the United States for the first time in 2024, placing China first globally. The emergence of DeepSeek has also shifted the tenor of China-U.S. competition in large AI models. Although still a follower, China’s advance is steady and forceful.

On 12 September, the State Council reviewed and adopted the Draft Regulations on the Administration of Clinical Research and the Clinical Translational Application of New Biomedical Technologies, signalling that new biomedical technologies will reach clinical use more rapidly and deliver wider public benefit. If China resolves to break down information silos and build high-quality biological datasets; to dismantle administrative barriers and fully leverage the full life cycle of a 1.4-billion population—the world’s largest application scenario; to break the spell of short-term capital exits so more patient capital focuses on this long-term track; and to end the separation of industry, academia, and research so that enterprises become true principal innovators, China’s AI+BT will no doubt stand at the global forefront.

Finally, I want to share an inspiring vision of AI, captured in a vivid line from a scholar: “The temple of science in the future may no longer be lit solely by the light of human wisdom. Across that vast domain of knowledge, human intuition, curiosity, and deep insight will shine alongside the tireless computation, impartial logic, and formidable creativity of artificial intelligence, composing a magnificent chapter in the exploration of the mysteries of the universe and of life.”

Big Science, Chinese Style: BGI’s Mei Yonghong on Reshaping the Research Paradigm from Govt Grant-guzzling to Profit-making Businesses