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

Govt official-turned-biotech corporate leader says fragmented academic silos can no longer deliver breakthroughs. Instead, he champions enterprise-led research, systems integration, and "know-how"

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, an influential Georgetown University thinktank under the Joe Biden administration, 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).

In Mei’s telling, China’s problem is not a shortage of clever people or cash, but an excess of academic silos and professors more adept at spending grants than solving problems. The traditional research paradigm of fragmented, curiosity-driven projects—the lofty “know-what”—is in dire need of replacement. What China needs is an emphasis on that grinding accumulation of tacit “know-how,” which industrialises and streamlines the entire research process, and which generates real machines, fertilisers, or gene sequencers. In China’s new scientific order that Mei envisages, enterprises, not academics, must take the lead.

The article was published in 《世界科学》 World Science magazine, Issue No. 7, 2025, based on Mei’s speech at the Science of Science Shanghai Forum, Pujiang Innovation Forum, founded by the MOST and the Shanghai Municipal government. It was reposted on the official WeChat blog of Huagu Biotechnology and Bioindustry Research Institute, a non-governmental, non-profit think tank set up in 2023 by biotech and agricultural companies in China, including BGI’s affiliates.

大科学时代：科学范式正在被重塑——兼谈大目标下的科技路径选择

The Era of Big Science: The Reshaping of the Scientific Paradigm—And on Choosing Technological Pathways Under Grand Objectives

BGI Group is composed largely of young people, with an average age just over 30. Many of its front-line researchers have only recently graduated. Yet BGI’s scientific output ranks among the very best in China’s life sciences. To date, it has published more than 730 papers in the leading journals Cell, Nature, and Science, including over 180 in the flagship editions. How is it that a private enterprise that does not treat paper publication as its primary goal can achieve such remarkable results? I believe the answer lies above all in a shift of research paradigm.

At present, most research institutions still follow the Principal Investigator (PI) paradigm: individual scientists or small teams pursue curiosity-driven inquiries, seeking to explain entire systems by studying their fundamental constituents. This paradigm of “small science,” rooted in reductionism, has dominated academic research for centuries, with towering figures such as Isaac Newton and Albert Einstein as its archetypes.

Since joining the International Human Genome Project thirty years ago, however, BGI has pursued a path distinct from reductionism: one of large platforms, big data, and broad collaboration. Its projects are all problem-oriented, and with big data as their foundation, the entire research process is streamlined and engineered, integrating multiple disciplines and uniting science, technology, and engineering. In today’s era of big data, BGI’s paradigm of “big science” clearly enjoys unique advantages.

In fact, this is not a coincidence but part of a broader trend. Long before World War II, the paradigm of “big science” had already taken shape in the United States, exemplified by the Manhattan Project and the Apollo moon-landing program. This new model dismantled disciplinary boundaries, fostered genuine interdisciplinary integration, and, above all, moved academia beyond the traditional ivory tower by aligning it closely with economic and social needs. To a great extent, the reason human civilisation has achieved more scientific and technological progress in the past century than in the previous two millennia combined lies precisely in this shift.

Consider the lithography machine, a critical “bottleneck” in China’s semiconductor industry. Where exactly does the bottleneck lie? Not in fundamental theory, but in engineering. A single lithography machine is made up of 100,000 components supplied by more than 5,000 vendors worldwide, requiring collaboration across dozens of countries. Some, therefore, argue that it is nearly impossible for any one nation to manufacture such a machine on its own.

Looking at ASML’s development of lithography technology, its progress was built on two decades of R&D at Philips and has accumulated over sixty years of continuous advancement. This is a mega-engineering achievement, requiring factories worldwide to etch hundreds of billions of transistors per second with precision finer than one-thousandth the width of a human hair. The process involves exacting demands across time, accuracy, speed, cost, strength, stability, yield, and more. The entire lithography industry has progressed through continuous trial and error and iterative engineering, driven by the construction of a global supply chain, the innovation of business models, and a problem-oriented response to market demands.

Anyone who has studied the basic theory of semiconductors understands this. Is the issue a shortage of talent, funding, market demand, or policy support? Clearly not. South Korea and Taiwan are not necessarily superior to the Chinese mainland in these respects. So what, then, is lacking? In my view, what the Chinese mainland lacks above all is engineering iteration and the tacit knowledge and experiential insight that accumulate from it. The scholastic model of compartmentalised disciplines, together with the fragmented, individualised Principal Investigator (PI) paradigm, is incapable of addressing these challenges.

Therefore, big science, as I understand it, can no longer be confined to the oft-cited notion of moving “from 0 to 1.” The journey “from 1 to 100” is just as much a part of scientific discovery and the accumulation of knowledge. Human knowledge may be divided into explicit and tacit forms. If “0 to 1” represents explicit knowledge, then “1 to 100” embodies tacit knowledge—implicit and unstructured. By analogy with an iceberg, explicit knowledge is the visible tip above the water, while tacit knowledge is the vast mass hidden beneath. Acquiring tacit knowledge requires interdisciplinary integration, through which products are redefined. In this sense, the complexity of moving “from 1 to 100” far surpasses that of “0 to 1,” and it can only be achieved within the paradigm of big science.

Recently, Fei-Fei Li, known as the “godmother of AI” in Silicon Valley, observed that the landscape of AI research has fundamentally changed: “Academia no longer possesses most of the resources in AI,” whether in chips, computing power, or data. These challenges are now more readily addressed in the marketplace and industry. Nature has likewise reported that many scientists are moving from academia to industry, where they enjoy much higher job satisfaction. The emphasis on practical applications drives more reliable research outcomes, while researchers themselves are better able to sustain patience, confidence, and curiosity. Underlying this structural shift is a transformation in the very paradigm of scientific research.

I would like to touch on a few related issues.

The first concerns the “reshaping of science.” Before the seventeenth century, science was largely the pursuit of individuals or small schools engaged in free inquiry. By the eighteenth century, it took the form of loosely organised academies. The nineteenth century saw the rise of collective models, and by the twentieth century, science had expanded to national and even international scales, with integration and technologization gradually becoming the dominant trend.

Song Jian, a former director of China’s State Scientific and Technological Commission and a strategic scientist, has argued in his discussion of systems theory and reductionism that reductionism is constrained by inherent theoretical limitations—chief among them its neglect of information, which, alongside matter and energy, constitutes one of the universe’s “three fundamental categories.” While reductionism portrays natural processes as reversible, biological evolution is irreversible, and time does not flow backwards.

Nobel laureate Professor Tsung-Dao Lee once made a similar observation: “To think that understanding [microscopic] elementary particles is equivalent to understanding the [macroscopic] vacuum is mistaken. Viewed simplistically, there would be no dark matter, nor phenomena such as quasars. The same is true of genetic organisation; knowing individual genes one by one does not unravel the mystery of life. Life is macroscopic.”

For the past thirty years, BGI’s research has consistently followed the path of holism and systems theory, in contrast to the mainstream trajectory of genetic research. For example, twenty years ago, when BGI committed to joining the International Human Genome Project, the academic mainstream was sceptical. Many dismissed those “dry numbers” as useless, preferring instead to devote their efforts and resources to functional genomics, seeking to analyse and exploit specific functional genes. The problem, however, is that life is complex—a mere aggregation of genes cannot account for a complete living organism. If the structure itself is not understood, how can one hope to master or regulate functional genes? Years later, many functional genomics studies ended without a conclusion, precisely because their underlying logic was flawed.

The second point concerns the role of the scientific community. In the era of big science, the value of the collective far outweighs that of individual contributions, as the age of scientific heroes quietly yields to collaboration and division of labour. The rise of AI and large language models has further overturned traditional research paradigms, rendering many lines of research in the life sciences obsolete. This new paradigm calls on scientists to step out of the ivory tower: to respect academia, but not be confined by it, and to engage actively with industry. At the same time, research in the era of big science will no longer be an elite pursuit but a vast, integrative system. With the arrival of the AI era, science has become open to all. In the future, increasing numbers of “blue-collar” scientists may well emerge as the main force driving the discovery of new knowledge and the creation of new industries.

The third point concerns “path selection for China’s science and technology.” I would like to offer six personal reflections.

—On the “top-level design” of the new system for mobilising resources nationwide. The United States hosts many highly influential universities, with seven of the world’s top ten universities located there. Yet even so, it relies on national laboratories and federally funded research institutes to pursue goal-oriented national projects. Such research is largely mission-driven and effectively supports the country’s strategic needs. This constitutes the American-style nationwide mobilisation system, characterised by clear top-level design, nationwide coordination, civil–military fusion, and the unification of industry, academia, and research.

China enjoys even greater institutional advantages, which makes it all the more essential to employ top-level design to comprehensively address strategic, directional, and overarching issues in science and technology. At the same time, it is crucial to overcome fragmentation across government departments, regions, civil and military sectors, and the spheres of industry, academia, research, and application. Only then can coordinated development be advanced in setting major directions, advancing major projects, formulating key policies, and carrying out important reforms.

—On national laboratories being oriented toward national goals. In recent years, China has invested heavily in building national laboratories, with funding often amounting to billions or even tens of billions of yuan. Such an investment scale was unimaginable in the past and remains rare worldwide. How, then, can national laboratories set themselves apart from universities, research institutes, and enterprises?

In my view, national laboratories should, first, function as public platforms that provide goods and services for the whole society. Second, they must be driven by clear national objectives, concentrating on the resolution of critical problems. Third, they should break down disciplinary silos, foster interdisciplinary collaboration, and pursue the systematic integration of science, technology, and engineering. Fourth, they should maintain full openness, linking industry, academia, research, and application, while actively engaging enterprises and a wide range of societal stakeholders.

Institutional innovation in national laboratories is essential. It is important to avoid simply “putting new wine into old bottles.” Where objectives are not being met, timely adjustments must be made. Scientific research activities should not come with an “iron rice bowl.”

—On the market-oriented self-sustaining value loop, where value circulates and interacts rather than being supplied unilaterally. Curiosity-driven research should never be dismissed, especially when it addresses fundamental scientific questions, for curiosity carries an irreplaceable and unique value. At the same time, it must not crowd out another essential driving force: market orientation. As Engels observed a century ago, “If society has a technical need, that helps science forward more than ten universities.”

At present, low efficiency in the allocation of innovation resources remains a major challenge in China. Although the Chinese government has invested trillions of yuan in science and technology, total factor productivity is still only about 40 per cent of that in countries such as the United States and Japan. A large share of scientific resources and R&D activity remains disconnected from economic activity, forming closed academic loops that run from project approval to completion.

The key solution lies in fully leveraging market mechanisms to break down these isolated loops and enhance the efficiency of resource allocation. Only by leveraging the role of the market can innovation gain a continuous internal driving force.

I believe, first, that enterprises should take the lead in industry–university–research collaboration. Being closest to the market, entrepreneurs best understand competition, and only by allowing enterprises to lead such collaboration can innovation elements be allocated across the entire chain to realise their full value. Second, intellectual property rights must be rigorously protected. Patents are private rights, and their protection is a fundamental requirement for building an innovation-driven nation; this must be comprehensively strengthened and enforced at the legal level. Third, all types of enterprises should be treated equally, with particular emphasis on providing stable institutional safeguards for private firms. Without enterprises as the principal actors, innovation cannot take place.

—On the principle of “equal treatment,” with enterprises as the main actors. Profound changes have taken place in the structure of science and technology in China: enterprises are no longer merely supplementary, but increasingly a vital force, and in some cutting-edge fields, even the dominant player. Yet serious misallocations remain in the distribution of scientific and technological resources. In the United States, for example, 80 per cent of Ph.D. graduates enter enterprises, whereas in China, 80 per cent go to universities and research institutes. Given that corporate R&D investment already accounts for more than 70 per cent of the national total, why is top-tier talent still so poorly allocated?

There is a strange phenomenon nowadays: Some government departments and officials seem to particularly favour institutions that are good at asking for and spending money, while showing little interest in businesses that generate profits, even keeping them at arm’s length or discriminating against them. Similarly, in the scientific community, the most influential are often academic elites skilled at spending funds, rather than enterprises that earn their own money and solve real-world problems. It’s time for a change. Enterprise-led collaboration between industry, academia, and research, with equal treatment for all innovation entities, is the only way forward.

—On the spirit of “daring to be the first down untrodden paths” in science. For many years, path dependency has posed a persistent challenge. There is nothing wrong with learning from the advanced, but this should never mean trailing passively in their footsteps without the courage to strike out beyond them. Moreover, China has already taken the lead in many fields, raising a critical question: how far can the original path still carry us?

The renowned “Qian Xuesen Question” not only asked why China has failed to produce first-rate talent but also went further: “Why do we lack the courage to be the first?”

[The Qian Xuesen Question (“Why has China failed to produce first-rate talent?”) has become a powerful motif in China’s public discourse on higher education. Interestingly, Qian himself, a U.S.-educated rocket scientist and the “father of China’s missile and space programme,” never posed the question in precisely that form. What he did say was: “One important reason China has not yet fully developed is that not a single university has been run with the model of cultivating talent in scientific and technological invention and creation. There is nothing truly original or innovative, and as a result, outstanding talent never emerges. This is a very serious problem.” After his death, however, the distilled version of “Qian Xuesen Question” quickly became a widespread public concern. —Yuxuan’s note]

For example, in the field of artificial intelligence. After the emergence of DeepSeek, some hailed it as a breakthrough capable of changing “the nation’s destiny.” When the United States restricted access to GPUs and computational power seemed an insurmountable barrier, falling behind appeared almost inevitable. Yet DeepSeek’s success has demonstrated that all roads lead to Rome—there is more than one path to scientific achievement. By drawing on distinctive characteristics, strengths, and needs to chart the right course, it is possible to reach the summit more effectively and swiftly.

Another example is healthcare. The United States, with a population of 340 million, spends $4.6 trillion annually on healthcare. China, with 1.4 billion people, would need to allocate its entire GDP to match the U.S. per capita spending level, yet still fall short of America’s current scale. A human life, from fertilisation to death, spans roughly a century. Why not make lower-cost prevention the foremost priority?

BGI launched a public welfare project in Hebei—newborn genetic screening—with a cost-benefit ratio of 1:17. In other words, every 1 yuan invested saves 17 yuan in public expenditure. What other form of public spending could possibly deliver such an extraordinary return on investment?

Here’s another example from modern agriculture. At present, China channels vast resources into seed breeding—it has the largest number of researchers in this field worldwide, and government investment is also the highest. Yet what has been the outcome? China’s seed industry still carries minimal weight in the global market. The real “bottleneck” is not seed research itself. Like humans, crops undergo a full life cycle from seed to harvest, and factors such as water, fertilisers, air, soil, and plant protection are all crucial to their growth.

While breeding is certainly important, it is by no means the only factor, and in many cases, it is no longer even the most decisive one. Over-concentration of resources on breeding ultimately caters more to academic preferences. By contrast, a microbial fertiliser developed by BGI costs only 5 yuan per mu [6.07 acres] yet can raise soybean yields by more than 10 per cent.

—On the bioeconomy as a matter of “the nation’s destiny.” Over the past three decades, life sciences and technologies have advanced at a breathtaking pace, yielding breakthroughs such as genetic modification, gene editing, synthetic biology, and biomanufacturing. Looking ahead, still more revolutionary biotechnologies may emerge, including DNA digital data storage, CO₂-to-carbohydrate synthesis, and artificial wombs. These developments would amount to a new scientific and technological revolution, one with unprecedented implications for humanity.

For China, this represents a historic opportunity tied to the nation’s destiny. In many critical fields, the country now stands at a starting line not far from that of the developed world. By leveraging its unique advantages in talent, big data, and application scenarios, China has the potential to make the bioeconomy a driving force in cultivating new quality productive forces and a cornerstone of the great rejuvenation of the Chinese nation.