Exclusive Interview | Understanding and Developing New Forms of Smart Economy

Editor’s Note: Artificial intelligence (AI) is penetrating all sectors of production and spheres of daily life, emerging as a core driving force for industrial upgrading and public wellbeing improvement. This year’s Chinese government work report for the first time called for the development of new forms of the smart economy and laid out important plans for advancing this agenda. To deepen understanding of what this concept entails and how it can be brought to fruition, a Qiushi reporter conducted an exclusive interview with Liu Dongmei, a research fellow at the Chinese Academy of Science and Technology for Development, Zhang Hui, a professor at Peking University, and Ma Yuan, a research fellow at the Development Research Center of the State Council.

Reporter: From the AI Plus Initiative to new forms of the smart economy, China is reshaping the pathways toward high-quality economic development by empowering various sectors with AI. How should we understand the smart economy, and what are the strategic considerations behind its development?

Liu Dongmei: The smart economy is an AI-driven economic form under which data is a key production factor and computing power and algorithms are important production tools. Its basic features are intelligent forms of production organization, business models, and consumption patterns. By means of human-machine collaboration, cross-sector integration, and co-creation and sharing, it enables the development of more intelligent, networked, and self-evolving economic activities. By using intelligent technologies to restructure the full chain of production, distribution, exchange, and consumption, the smart economy enhances the efficiency of resource allocation, stimulates innovation, and transforms industries and broader economic ecosystems. Fundamentally, the smart economy is the natural result of the social productive forces reaching a new stage of development.

Visitors tour the computing infrastructure exhibit area. The experience zone at the 9th Digital China Summit opened to the public at the Fuzhou Strait International Conference and Exhibition Center in Fujian Province, April 28, 2026. PHOTO BY CNS REPORTER LV MING

Zhang Hui: From the perspective of economic history, the smart economy represents a new economic form following the agricultural, industrial, information, and digital economies. Far more than a mere conceptual shift, its emergence represents a fundamental transformation in the logic of development. Historically, the core production factors of the agricultural economy were land and labor; the industrial economy was powered by capital and energy; and the information economy relied on the flow and interconnection of information. Building on these foundations, the digital economy has established data as a key factor of production. By completing the evolution from informatization to comprehensive networked development, it has laid a firm foundation for a more advanced economic form to emerge.

The smart economy is rooted in the convergence of three factors. The first is the explosion of data resources. The sheer density of information now exceeds the limits of human processing, necessitating new analytical tools. The second is a revolutionary breakthrough in computing power and algorithms, which has arisen due to the continuous improvement of computing infrastructure and rapid iteration of algorithmic models. This has enabled real-time processing of vast data volumes and facilitated autonomous decision-making. Third, to navigate highly complex market environments, industries must move beyond simply connecting information to interpreting it and making predictions. Driven by both the intrinsic needs of economic development and the cumulative effects of technological iteration, the economy is making a leap from being data-driven to intelligence-driven. The rise of the smart economy does not represent the emergence of something entirely new. Rather, it signifies a major advance in the evolution of the digital economy from connectivity to intelligence and from empowerment to reconstruction.

Ma Yuan: At present, with global competition in science, technology, and industry intensifying, countries are stepping up efforts to build AI capabilities. The smart economy has become a critical arena of competition among major countries. On the one hand, developing the smart economy is imperative for securing new advantages in global competition. Success in this regard hinges on firmly securing the initiative in underlying technologies, data resources, and application ecosystems. To position itself successfully to prevail in this intensifying global competition, China must make the right moves early, putting in place a comprehensive strategy for AI that encompasses both technological innovation and governance frameworks. On the other hand, the smart economy is a fundamental force for cultivating new quality productive forces and unleashing new growth drivers. China has already made initial progress in building large clusters in emerging industries such as AI computing chips, large models, and embodied intelligence. The growth of these industries is catalyzing the rapid advancement of future industries such as brain-computer interfaces, quantum computing, and 6G. By creating new demand through innovations in supply, AI is simultaneously attracting substantial investment and fostering new areas of consumption, making it a key driver of higher factor productivity and building a robust domestic market.

Reporter: Developing new forms of the smart economy is both an inevitable outcome of technological advancement and a proactive strategic choice for national development. In what key respects can these emerging forms be considered “new” ? Compared with traditional models of economic growth, what new opportunities will their development create?

Liu Dongmei: At the level of the workforce, workers specializing in the application of intelligent technology have become a crucial force driving the development of the smart economy. Human-machine collaboration—the close integration of the human brain with intelligent computing—is poised to become a standard feature of new labor models. At the level of the means of labor, conventional mechanized and automated equipment is giving way to large AI agents with robust autonomous learning and reasoning capabilities. At the level of the objects of labor, vast stores of data generated through the internet and proprietary data drawn from real-world settings have become new means of production capable of being extensively processed. At the organizational level, production is no longer confined to linear workflows within a single entity. Intelligent technologies can now be leveraged to achieve networked collaboration across entities, domains, and regions, substantially improving the precision, real-time responsiveness, and autonomy of production processes.

Ma Yuan: Specifically, the new character of these emerging forms of the smart economy is evident in the following respects. First, the technological architecture is new. AI is advancing from small to large models, from single-modality to cross-modality, and from specialized to general intelligence. Continued breakthroughs across these areas are unlocking unprecedented development opportunities. Second, new infrastructure is required. The training and deployment of large models demand enormous computing power. This calls for forward-looking planning to develop hyper-scale intelligent computing clusters, along with simultaneous steps to optimize the planning, expansion, and upgrading of supporting energy and network infrastructure. Third, new production models are emerging. Powered by AI, manufacturing is transitioning from traditional automated assembly lines to flexible and collaborative production models, enabling manufacturers to respond more efficiently to demand for a wider variety of customized products in smaller batches. Fourth, new work tasks are being created. The development of AI has given rise to an unprecedented range of tasks, including data governance, model training, system testing, algorithm deployment, operations and performance optimization, and ethics governance, all of which are creating many new employment roles. Fifth, a new relationship between humans and machines is taking shape. No longer limited solely to manual labor, machines are increasingly undertaking complex cognitive tasks. In the process, they are gradually evolving from tools into partners in the workplace. This is driving a shift from traditional human-to-human collaboration to a new model of human-machine collaboration.

The smart economy is reshaping the underlying logic of economic operations. Growth is becoming less based on scale and more driven by cognitive capabilities, thereby opening up new opportunities for expansion. One example is the increasing momentum of the large model industry. A new generation of large models, equipped with multimodal understanding, complex reasoning, and agentic capabilities, is being deeply integrated into terminal devices such as smartphones and computers, giving rise to a new wave of AI-native enterprises. Another example is the rapid expansion of the intelligent computing industry. In the first quarter of this year, China’s intelligent computing capacity reached 1,882 EFLOPS (exa floating-point operations per second), placing it among the world’s leaders. This is indicative of the strong pull on infrastructure development from the surge in AI applications. It is laying a solid foundation for the long-term development of the computing power industry. Furthermore, the embodied intelligence industry has become an investment and financing hotspot. By the end of 2025, financing in China’s embodied intelligence and robotics sectors exceeded 70 billion yuan. There is strong potential for these technologies to expand into areas such as industrial manufacturing and consumer services, reshaping the industrial ecosystem of human-machine interaction in the process. In addition, AI is producing significant gains in productivity. At the end of 2025, the key process numerical control rate in enterprises across China’s key industries reached 68.6%. As AI applications are being rapidly rolled out across finance, manufacturing, and other industries, they are delivering notable empowerment outcomes and demonstrating considerable potential to boost quality and efficiency.

Reporter: China’s smart economy is showing strong momentum as technological breakthroughs are translated into large-scale applications. What advantages and conditions does China have to draw on in developing the smart economy? What prominent issues and challenges must it confront?

Zhang Hui: When discussing China’s advantages in developing the smart economy, it is important to focus not just on technology itself, but also on the conditions that underpin its growth. First, there is a strong connection between China’s application scenarios and its markets. With a population of more than 1.4 billion, the world’s largest middle-income group, and a comprehensive industrial system, China offers a wide range of complex application scenarios spanning manufacturing, agriculture, and public governance. This is what has made it the most dynamic “natural testing ground” for intelligent technologies in the world, where new advances can move rapidly from laboratory to large-scale deployment. Second, China possesses a solid infrastructure base. The country is a global leader in digital infrastructure, including 5G base stations, data centers, and the Internet of Things. In addition, new infrastructure, such as intelligent computing clusters, is being rapidly deployed, while computing capacity and electricity supply are being developed in great coordination, providing strong support for cost-effective technology applications. Third, China benefits from a continuous supply of vast data resources. During the previous wave of digitalization, the country accumulated vast amounts of valuable data across the government, industry, and consumer sectors. These data provide abundant fuel for the iterative upgrading of the smart economy. Fourth, China possesses notable institutional strengths. Over the years, the country has introduced a series of national strategies and policies, creating a comprehensive policy framework covering the entire chain of technological innovation, industrial application, infrastructure development, and security governance. By establishing a steady and well-defined pathway for smart economy development, this framework has helped foster a virtuous cycle where technological breakthroughs and industrial applications reinforce each other via hands-on learning and application-oriented innovation.

Workers operate on an intelligent production line at a “future factory” run by an enterprise in Lvshan Township, Changxing County, Huzhou City, Zhejiang Province, March 16, 2026. XINHUA / PHOTO BY TAN YUNFENG

Ma Yuan: As a fundamental engine steering future economic development, the smart economy will become a direct determinant of a country’s position in global industrial, innovation, and value chains. Beyond the need for technological innovation and catch-up, the core issue for China at present lies in a structural mismatch between the supply of technology and market demand.

On the supply side, the foremost challenge is a shortage of high-quality data. China’s vast quantities of data are often cited as an advantage for the country. However, these data are not necessarily suitable for accurate labeling, cross-institutional sharing, or model training in compliance with laws and regulations. In addition, the institutional frameworks for data rights confirmation, circulation, and pricing also require further improvement. Second, intelligent computing capacity remains relatively constrained and has not kept pace with the explosive growth in inference and training demands. Third, large-model technologies are not yet fully mature. Risks such as “model hallucination” (where models generate plausible-sounding but factually incorrect content) and “model poisoning” (where malicious data are used to interfere with training, resulting in distorted or biased outputs) highlight weaknesses in trusted data sources and algorithm governance.

Looking at the demand side, enterprise clients are eager to use AI technologies to cut costs and boost efficiency. Yet they remain concerned about actual model performance, data security risks, and return on investment. At the same time, enterprises, especially small and medium-sized ones, differ in their levels of digital maturity. Many lack cross-disciplinary professionals with expertise in both AI and their specific fields. As a result, they often find themselves wanting to adopt these technologies but are wary of doing so, unsure how to use them, or unable to afford them.

Liu Dongmei: At a deeper level, the structural misalignment constraining the development of the smart economy is fundamentally a question of how to improve the entire system. This means that weaknesses in the smart economy rarely exist in isolation; they arise from the interplay of multiple technological, institutional, industrial, and governance challenges. On the technological front, ongoing shifts in global competition have steadily raised the barriers to acquiring high-end chips, foundational software, and advanced manufacturing equipment. China also continues to face weaknesses in key areas such as algorithms, large models, and operating systems. Besides, its capacity for technological self-reliance and control needs to be further strengthened. In terms of applications, bottlenecks are hindering the full integration of AI with the real economy. The commercialization pathways for intelligent technologies remain under exploration, and in some areas, technologies have yet to be effectively matched to application scenarios. Truly replicable, scalable benchmark cases and mature models have yet to emerge. From a governance perspective, as AI technologies penetrate various sectors with growing speed, issues such as algorithmic ethics, data security, and structural employment shifts are becoming increasingly pronounced. Existing governance frameworks have not kept pace with the development of the smart economy and are ill-adapted to its needs. Furthermore, the contest to shape international rules cannot be overlooked. Challenges such as trade and economic frictions, technological competition, and the reconfiguration of global industrial and supply chains have a direct bearing on the high-end production factors, international markets, and rule-making space that China needs to develop its smart economy.

Reporter: This year’s government work report sets out systematic arrangements to foster a new form of the smart economy, charting a clear roadmap for its high-quality development. What key practical requirements should guide this process?

Ma Yuan: Fostering new forms of the smart economy is a long-term, systemic endeavor. At present, with AI technologies evolving at a rapid pace, their development must be kept firmly on track and advanced in a well-managed manner. First, equal emphasis should be placed on both technological breakthroughs and practical application. We should support research on original and foundational technologies while accelerating the deployment of mature AI technologies across all industries and sectors. Second, category-based policies and targeted guidance should be pursued in tandem. We must identify the respective demands and institutional barriers facing government institutions, businesses undergoing transformation, and the general public and provide support accordingly. Third, breakthroughs in key areas should be pursued in tandem with systematic advancement. We must respect the inherent logic underlying technological progress and the respective stages through which AI becomes integrated with the real economy. We must give priority to industries with solid foundations and robust development potential, and refrain from reckless advancement and blind expansion in disregard of real conditions. We must also avoid wasting resources that can emerge from undifferentiated competition. Fourth, demand pull and supply push should work in tandem and be aligned with the actual capabilities of AI models, to avoid mismatches between supply and demand that are divorced from technological reality. Fifth, development, governance, and security should be coordinated. Pilot programs should be carried out to identify risks and address institutional weaknesses, and faster steps should be taken to establish an institutional framework tailored to the smart economy.

Zhang Hui: The smart economy is more than a simple combination of computing power, algorithms, data, applications, and governance. Rather, it is an integrated whole where each dimension mutually interacts with and defines the others. In practice, several dimensions warrant particular attention.

When building infrastructure, rather than focusing on scale alone, we should carry out forward-looking assessments to establish how decisions will affect the application and governance layers. For example, new infrastructure projects, such as hyper-scale intelligent computing clusters and coordinated computing capacity and electricity supply, must be designed alongside mechanisms for data circulation and utilization. Otherwise, if high-quality data are unavailable or cannot circulate freely, the effectiveness of installed computing capacity will be severely undermined.

When fostering the industrial ecosystem, technological bottlenecks should not be tackled in isolation. Breakthroughs in chips, algorithms, and large models tend to depend on vibrant open-source AI communities and high-quality datasets. As such, only by turning the smart economy from a game of a few tech giants into a toolkit for small and medium-sized enterprises can we foster a flatter, more dynamic market, rather than entrenching monopolies.

When deploying applications and establishing industry standards, we should carry out systematic planning. The rollout of intelligent terminals and AI agents depends on simultaneous improvements to the rules and systems covering ethics, security, and regulation. Only by advancing development, governance, and security within a unified framework can we unlock the smart economy’s intrinsic value and avoid oscillating between unchecked growth and excessive regulation.

Reporter: Developing new forms of the smart economy requires further refinement of institutions and mechanisms to foster relations of production better aligned with such new forms. At present, what key areas of reform should be given priority to ensure that production relations better adapt to the smart economy’s ongoing development?

Zhang Hui: It is a fundamental law of social development that the relations of production must evolve in step with the development needs of productive forces. In previous technological revolutions, tools were in essence used to extend human capabilities. At their core, the relations of production centered on collaboration and distribution among people. In the age of the smart economy, however, AI agents cannot only perform specific tasks in place of humans, but also participate deeply in core processes such as decision-making, collaboration, and value creation. This is giving rise to a new dynamic where humans and AI agents both compete and cooperate with each other. Therefore, as new forms of the smart economy trigger the systematic restructuring of production modes, they will inevitably necessitate the formation of new relations of production better aligned to its development. This issue deserves close attention.

Liu Dongmei: Fostering new forms of the smart economy represents both a development imperative and a reform task. We must strive to remove institutional barriers constraining the development of the smart economy and channel various advanced production factors into these new forms.

First, we must move faster to refine the market-based mechanisms for allocating data as a production factor. This means exploring and establishing a foundational institutional framework that balances efficiency and fairness in data rights confirmation, circulation and trading, income distribution, and security governance. Efforts should be made to break down information silos arising from entrenched interests and address the ongoing hesitation and reluctance to share data, so that data resources can be converted into assets and capital. This will help consolidate the foundation of production factors underpinning the smart economy’s development.

Second, we must ensure patient capital lends stronger support for innovation and entrepreneurship in the smart economy. To cultivate patient capital for smart industry ventures, we must deepen reform of the financial and investment systems. We should leverage the role of government guidance funds and improve the risk-sharing and exit mechanisms for investment and financing in key smart economy sectors. These steps will help ensure a steady flow of capital to support innovation and entrepreneurship in the smart economy.

Third, we must accelerate the establishment of an industrial regulatory framework tailored to the development of the smart economy. Relevant market access restrictions should be eased in an orderly manner, and further efforts should be made to remove administrative monopolies and market barriers to ensure fair competition among market entities of all types. We should also create new regulatory approaches to enable the piloting and testing of new technologies and business models in selected sectors and within clearly defined parameters.

Finally, we must steadily refine our employment and income distribution systems. Forward-looking steps should be taken to establish mechanisms that promote employment opportunities based on human-machine collaboration, and occupational standards and skill certification systems should be refined to cater for new and emerging roles. We should also explore mechanisms to effectively recognize data as a production factor in income distribution, ensuring that data contributors and algorithm developers receive a fair share of the value-added benefits generated by the smart economy.

(Originally appeared in Qiushi Journal, Chinese edition, No. 10, 2026)