Originality Is Not a Matter of National Intelligence, but of Social Systems

China’s high-speed rail, skyscrapers, e-commerce, mobile payments, solar industry, electric vehicles and AI applications are often treated as straightforward proof of modern capability. Those achievements are real and substantial. Yet a deeper look reveals a different picture. Much of the basic theory, tooling, engineering vocabulary and organisational method of modern technological civilisation was formed in Western systems, above all in the United States.

This does not mean that China lacks engineers or can do only low-level assembly. On the contrary, it has a vast number of capable and adaptable engineers, together with formidable strengths in mobilisation, supply-chain integration, cost control and deployment at scale. But rapidly enlarging, cheapening and popularising an existing technology is not the same capability as continually producing the next technological paradigm, engineering language or framework of knowledge.

After several decades of software development work, this is particularly clear to me. Almost the entire working vocabulary of a contemporary software engineer — computer architecture, operating systems, databases, network protocols, programming languages, object-oriented and functional programming, modularity, design patterns, version control, automated testing, continuous integration, continuous delivery, cloud computing, containers, microservices, Agile, DevOps, SRE, and lifecycle management from requirements to maintenance — was not first formed in China.

Even when these ideas are reduced to fashionable labels such as waterfall, Agile, Scrum or DevOps, they stand for more than a few slogans. They embody accumulated engineering constraints: how requirements change, where responsibilities lie, how quality is traced, how failures are reviewed, how versions are rolled back, how deployment is automated, how user feedback returns to the next development cycle, and how different teams cooperate under common rules. The distinctive Western strength has not merely been the production of individual software products. It has been the repeated conversion of experience into methods that can be taught, tested, reused and revised.

China’s software industry uses these methods, often with remarkable speed. But implementation frequently retains only the parts that accelerate immediate delivery, while removing the slower and less visible practices that protect quality and responsibility. Agile becomes an excuse for less documentation and faster releases. DevOps becomes a request for developers to take on more operations work. Project management becomes pressure on deadlines. This can look flexible and efficient in the short term, but it is often purchased at the cost of maintainability, testability and long-term accountability. What is called flexibility is sometimes simply the removal of constraints whose value is not immediately visible.

This helps explain why China can appear especially dynamic at the application layer. Chinese input methods, localised interfaces, payment and e-commerce flows, industry-specific business systems, cost reduction, platform operations, and the recombination of existing technologies into new services all leave considerable room for initiative. These are real innovations with practical value. Yet they mostly sit on top of pre-existing platforms, protocols, chips, operating systems, databases and cloud infrastructure. Combining applications creatively is not the same as redefining the technological foundations beneath them.

Artificial intelligence follows a similar pattern. Chinese researchers, engineers and companies participate deeply, and are strong in model training, open models, deployment, Chinese-language data, productisation and cost optimisation. But the principal intellectual route and infrastructure of modern AI — the revival of neural networks, deep-learning frameworks, the GPU computing ecosystem, the Transformer, large-scale pretraining, cloud-scale computing organisation and the contemporary frontier-model company — have been shaped mainly by European and American institutions. China can catch up rapidly along this track, and may surpass others in particular applications or cost structures. That does not mean it laid the track in the first place.

Still, it would be wrong to turn this into the claim that Chinese people lack creativity. Modern technological creation is rarely the achievement of one person sitting at a desk. It depends on enormous systems of researchers, engineers, universities, laboratories, suppliers, capital, law, open-source communities and international collaboration. In the modern sense, whether a nation has creative capacity is not primarily a judgement about racial ability or individual intelligence. It is a judgement about whether its society can reliably generate, protect, test and extend original work.

For that reason, the large presence of Chinese and Chinese-born researchers in American technology companies, universities and laboratories does not refute this argument. It clarifies it. Individual ability is not the decisive barrier. Put the same person into a different institutional and intellectual environment, and the available resources, acceptable risks, scope for criticism, ability to pursue apparently unproductive work, and means by which peers absorb and develop a result can all change radically. Individual creativity becomes durable public creativity only when a social system can receive it.

My own experience in Australia has often suggested that Chinese and other Asian migrants in technical roles may have stronger educational preparation and take on more technically demanding work than the local average. This is not evidence that Westerners are less intelligent, or that Asians are inherently more intelligent. Migration is itself selective: many skilled migrants arrive with strong educational and professional advantages, while many people in the local population do not make advanced technology the centre of their work. The observation shows, above all, why distributions of individual ability cannot substitute for an analysis of social institutions.

American creativity does not mean Americans are innately smarter. The United States has remained central to modern technology because it built an innovation system able to connect talent, capital, universities, defence demand, start-ups and global markets from around the world. It permits extensive failure, competition among rival ideas, and the possibility that a new idea can obtain resources and be tested before it receives universal approval. Its politics and society have serious flaws. Yet the openness and self-renewing capacity of this system remain central to its technological advantage.

The real question for China is not how to prove that it is already no worse than the West. It is whether it can build an environment that continually produces basic theory, foundational tools, original standards and new engineering methods. That requires more than investment and slogans. It requires respect for professional judgement, space for serious criticism and failure, protection for long-term research, less substitution of administrative targets for intellectual judgement, and a way for methods to become shared industry constraints through open testing.

China’s engineering capacity is already strong enough to reshape the physical world. But engineering scale does not automatically create source-level innovation. A society that can only learn, integrate and enlarge frameworks established elsewhere may become powerful and achieve striking results, yet it remains constrained by a technological world defined by others. The harder task is to become a society capable of continually proposing and sustaining new frameworks of its own.