Observation on the development of China's digital economy

Reporter Guo Jinhui compiled it according to public information

Editor's note

The vigorous development of the digital economy has become increasingly prominent in leading and supporting economic and social development. A number of institutions have recently released the development of digital technology innovation represented by artificial intelligence. Some institutions have also released reports on digital technology enabling industrial transformation and upgrading. The mainland's digital technology innovation capabilities have been continuously enhanced, and digital technology and the real economy have been deeply integrated, which will continue to strengthen, optimize and expand the mainland's digital economy.

The development of the artificial intelligence industry is accelerating

Reading tips

All walks of life are accelerating the transformation from "+ artificial intelligence" to "artificial intelligence +". Several institutions have released the progress of the AI industry. CCID Consulting Artificial Intelligence and Big Data Research Center analyzes the comprehensive strength of China's leading enterprises in the field of large models from five dimensions: core technical capabilities, market competitiveness, product innovation capabilities, market development potentials and ecological opening capabilities. Tencent Research Institute released the industry's first industry model research report, which provides an in-depth analysis of the development background, application progress, implementation methods, security and governance, and future development trends of industry models. China Mobile Research Institute released the industry's first white paper that systematically expounds the core design principles and key technologies of the ultra-10,000-card cluster, and introduces the technical research and practical experience of China Mobile and industry partners in the construction of the ultra-10,000-card cluster.

CCID Consulting Artificial Intelligence and Big Data Research Center:

The competitive landscape of China's large-scale model enterprises is the coexistence of diversification and specialization

According to the "Research Report on the Development of China's Large Model (2024)" (hereinafter referred to as the "Report") recently released by CCID Consulting Artificial Intelligence and Big Data Research Center, the competitive landscape of China's leading large model enterprises shows the characteristics of diversification and specialization.

In recent years, China's large-scale model technology has undergone rapid innovation and development, and the development of China's large-scale model technology has achieved remarkable results in many aspects. In terms of technological innovation, Chinese scientific research institutions and technology enterprises continue to invest in R&D resources, which has promoted the remarkable results of large models in algorithm optimization, architecture innovation, and training efficiency improvement. In terms of market progress, the application and commercialization process of China's large models is accelerating day by day, and practical cases have been formed in the fields of content creation, enterprise services, and government services. In terms of industrial ecology, universities and research institutes, as the source of research and development of basic theories and key technologies, work closely with technology giants, unicorn companies and start-ups to jointly promote the research and development, iteration and application of large-scale model technology.

The policy of the development of China's large-scale model industry focuses on innovation-driven and ecological optimization, which is embodied in the following aspects: integration of innovation system and strengthening of policy support, equal emphasis on data openness and sharing, privacy and security, cross-border integration and industry standardization construction, talent training and international cooperation acceleration, and promotion of the coordinated development of industrial ecology.

The development of China's large model highlights the dual emphasis on self-research capabilities and industry standardization, and the industry will develop in the following directions in the future. First, the core hardware is independent and controllable. In the face of the uncertainty of the international environment, accelerating the independent research and development of AI chips has become the cornerstone of the independent development of the industry. The second is the deepening of data governance and standardization. The industry will increase the standardization of data collection, processing and use, optimize the data circulation mechanism, and ensure the high-quality supply and compliant use of data. The third is flexible customization and micro-innovation strategy. Adaptively adjust on the basis of the general large model to meet the application of specific scenarios in a low-cost and high-efficiency way. Fourth, ethical norms and social responsibility have been strengthened. The industry will work together to explore the establishment of AI ethics guidelines to ensure that the development and application of large models are responsible for respecting privacy, maintaining fairness, and promoting social welfare.

Tencent Research Institute:

Industry models are expected to accelerate their application in traditional industries

According to the "Research Report on Industry Large Models" (hereinafter referred to as the "Report") released by Tencent Research Institute on May 13, under the guidance of important policies such as "artificial intelligence +", industry large models are expected to accelerate the application of traditional industries, and with the support of the infrastructure of cloud and intelligence, they will develop in the direction of multi-modality, artificial intelligence, end-side and miniaturization, and be more deeply embedded in the workflow of various industries, so as to promote the improvement of productivity.

The industry model as a whole is in the early stage of development, and there is no large-scale mature application example. Which industries are developing rapidly in the application of large models? According to the report, digital native industries (such as various Internet applications) are the pioneers, while producer services (advertising, finance, etc.) in traditional industries are making relatively fast progress, and asset-heavy industries (construction, manufacturing, energy, etc.) are making relatively slow progress. There are two core influencing factors behind it: the adaptability of demand and the availability of data, the higher the industry, the faster the progress.

According to the report, the development of artificial intelligence may be of greater significance to the industry's large model. Some artificial intelligence that is deeply coupled with the workflow has begun to emerge, and it is expected to gradually develop into a new type of productivity that is indispensable for all walks of life. For example, in the field of social networking, some AI products will further promote the intelligence of content creation and social interaction, bringing users a richer and more diverse experience. In the industrial sector, the application of industrial visual inspection robots shows a high degree of autonomy in identifying, judging and repairing equipment.

With the deepening of AI applications, the scale, type, and complexity of models will continue to increase, and Model-as-a-Service (MaaS) will increasingly become the mainstream way for industry users to use intelligence on the cloud. Users can either directly call the large model application programming interface (API) of the cloud or use the full lifecycle large model training tool to generate a large model suitable for their own scenarios and host it on the cloud to provide high-quality intelligent services for end users. To this end, a higher-performance computing power base for artificial intelligence is indispensable, and the efficiency of model training, development, and application can be comprehensively improved through optimization at all levels such as computing, storage, communication, and training.

China Mobile Research Institute:

The 10,000-calorie cluster has become the standard configuration of the large-scale model infrastructure competition

The White Paper on New Intelligent Computing Technology for Super 10,000 Card Clusters (2024) (hereinafter referred to as the White Paper) was officially released on April 29, led by a team of experts from China Mobile Research Institute. This is the industry's first white paper that systematically expounds the core design principles and key technologies of the Chaowanka cluster.

Since the release of ChatGPT, there has been a rush of competition for large models in the tech world. Data has become a new factor of production, computing power has become a new basic energy source, and large models have become a new production tool, and the transformation of all walks of life from "+AI" to "Al+" has become unstoppable. According to the white paper, as the number of model parameters moves from 100 billion to trillions, the model capabilities are more generalized, and the demands of large models for the underlying computing power are further upgraded, and the super 10,000 card cluster has become the standard configuration of this round of large model infrastructure arms race.

According to the white paper, the Chaowanka cluster will help compress the training time of large models, realize rapid iteration of model capabilities, and respond to market trends in a timely manner. However, how to achieve efficient training in a cluster of over 10,000 cards and maintain the stability of the training process for a long time is a dual challenge to scale large model training to tens of thousands of GPU cards. In the process of running the super 10,000 card cluster, key issues such as the effective computing power of the cluster, the stability guarantee of the ultra-large-scale Internet network, and the rapid troubleshooting and repair of faults are all the focus of attention in the industry.

The white paper comprehensively expounds the core technical requirements for the construction of ultra-10,000 card clusters from five aspects: cluster energy-efficient computing technology, high-performance converged storage technology, large-scale inter-server high-reliability network technology, high-fault tolerance and energy-efficient platform technology, and new intelligent computing room design, and provides technical suggestions for the construction of domestic intelligent computing infrastructure.

With the continuous expansion of data scale, the continuous enhancement of cluster capabilities, and the increasing abundance of large model applications, higher requirements are put forward for the upgrade of new intelligent computing bases. In the future, the white paper calls for technological breakthroughs in the fields of super-nodes, cross-cluster training, and software frameworks to strengthen the capabilities of intelligent computing infrastructure. At the same time, we will continue to explore the combination of advanced technology fields such as storage and computing integration and photonic chips with intelligent computing centers to lay the foundation for the next information revolution.

Digital intelligence promotes the "renewal" of the industry

Reading tips

Digital intelligence has brought unprecedented opportunities to many fields. A number of institutions have published studies in this regard. The research report led by the China Academy of Information and Communications Technology summarizes the typical paths of digital transformation of small and medium-sized enterprises and analyzes the key factors affecting the digital transformation of small and medium-sized enterprises. The research report released by the Institute of Financial Strategy of the Chinese Academy of Social Sciences analyzes the innovative practices and experiences of the digital economy to empower agriculture and promote rural revitalization, and points out that the digital economy empowers rural revitalization is a process of continuous exploration and gradual deepening. According to the research of Tianjin University of Traditional Chinese Medicine and other units, the development of digital intelligence of modern Chinese medicine is an inevitable trend to promote the innovation of traditional Chinese medicine, which summarizes the current situation of the development of digital intelligence of modern Chinese medicine, and puts forward the goals and action plans of the development of digital intelligence of modern Chinese medicine.

China Academy of Information and Communications Technology:

The digital transformation of small and medium-sized enterprises is still in a stage of rapid development

The Research Report on Digital Transformation of Small and Medium-sized Enterprises (2023) (hereinafter referred to as the "Report") compiled by the China Academy of Information and Communications Technology, telecom operators, other scientific research institutions, small and medium-sized enterprise software and hardware service providers and other relevant industry forces was recently released.

According to the report, the four main modes of promoting the digital transformation of small and medium-sized enterprises in mainland China are: First, the leading enterprise-driven model, in which industry leaders lead small and medium-sized enterprises in the same industry to achieve transformation by providing digital products and services. The second is the cluster and park-driven model, which promotes the collective transformation of SMEs in the region through the establishment of shared digital platforms and facilities. The third is the industrial Internet platform-driven model, which uses platform resources to drive the collaborative transformation of small and medium-sized enterprises and optimize supply chain management. Fourth, the "small, fast, light and accurate" technology product-driven model, that is, small and medium-sized enterprises adopt technology products that are easy to deploy and use to improve their operational efficiency and market response speed.

The transformation needs of small and medium-sized enterprises in different industries are different. Small and medium-sized enterprises in the raw material industry focus on improving the ability of safe production and research and development of new materials; Small and medium-sized enterprises in the equipment manufacturing industry focus on improving operational efficiency and research and development of precision components; Small and medium-sized enterprises in the electronic information industry pursue precise production control and product quality improvement; Small and medium-sized enterprises in the consumer goods sector are looking to become more responsive and profitable.

There are differences in the transformation paths of small and medium-sized enterprises of different sizes. The size of SMEs has a significant impact on their approach and approach to digital transformation. Medium-sized enterprises are able to implement complex digital applications across business processes due to their more complex business needs and sufficient resource capabilities. SMALL AND MICRO ENTERPRISES, ON THE OTHER HAND, PREFER TO USE STANDARDIZED SAAS SOFTWARE AND SIMPLE DIGITAL TOOLS TO DIGITALLY DEPLOY AROUND A SINGLE BUSINESS LINK.

According to the report, on the whole, the digital transformation of small and medium-sized enterprises will continue to develop rapidly in the coming period, and the development speed will continue to improve. In terms of application, the business field of digital transformation of small and medium-sized enterprises will further expand from general fields such as operation and management to special fields such as manufacturing, R&D and design, marketing and sales. In terms of technology, emerging technologies such as artificial intelligence will play an increasingly important role and value in the digital transformation of small and medium-sized enterprises.

Chinese Academy of Social Sciences:

The digital economy has effectively integrated and expanded the agricultural industry chain

The Chinese Academy of Social Sciences' Institute of Financial and Economic Strategy recently hosted a seminar on "Digital Economy Helps Rural Revitalization" and the press conference of the "Report on the High-quality Development of China's Digital Economy (2023)" (hereinafter referred to as the "Report"). According to the report, the digital economy has become an important driving force for rural revitalization, effectively integrating and expanding the agricultural industry chain, and continuously promoting industrial revitalization, talent revitalization, cultural revitalization, ecological revitalization and organizational revitalization through various ways.

In the process of promoting rural revitalization, the digital economy has gradually formed four models: the information platform model, the single product digitization model, the agricultural industrial park model and the three-industry integration model, and the new real economy is becoming an important subject of rural revitalization.

In the process of integrating the digital economy and the real economy, a series of new entities represented by JD.com and China Unicom have emerged in the mainland, which are the new subjects to help rural revitalization. New entity enterprises have special advantages in promoting technological innovation, actively reaching out to multiple market players, supporting agricultural operators to enter the digital ecosystem, providing application-oriented services, and cultivating digital talents.

In particular, important progress has been made in the digital transformation and upgrading of the industrial chain and supply chain, and the whole industrial chain service system has been gradually formed, providing good support for industrial revitalization. At the same time, the digital economy has also made positive progress in talent revitalization and rural governance.

The digital economy empowers rural revitalization thanks to targeted fiscal, tax and financial support policies, and the huge investment of financial funds and financial resources has laid a solid foundation for the development of the digital economy and rural construction. In the process of promoting the rural revitalization strategy, the country has carried out many innovative practices and achieved good results.

The report believes that the empowerment of rural revitalization by the digital economy is a process of continuous exploration, gradual deepening, and from success to greater success, and there are also some deficiencies, shortcomings or problems in development. In the future, it is necessary to use the digital economy to empower rural revitalization more scientifically, effectively coordinate various policies, elements and resources, and write a new chapter in the high-quality development of rural revitalization.

Tianjin University of Traditional Chinese Medicine and other institutions:

Accelerate the development of digital intelligence in modern Chinese medicine

To promote the development and practice of digital intelligence of modern Chinese medicine, under the guidance of dozens of academicians, masters of traditional Chinese medicine, industry experts, and the National Key Laboratory of Modern Chinese Medicine Creation, the China Association of Traditional Chinese Medicine, and the China Economic Information Service of Xinhua News Agency, Tasly Biomedical Industry Group Co., Ltd., Tianjin University of Traditional Chinese Medicine, Zhejiang University, Huawei Cloud Computing Technology Co., Ltd. and other units jointly issued the "White Paper on the Digital Intelligence Development of Digital Intelligence in Chinese Medicine-Modern Chinese Medicine" (hereinafter referred to as the "White Paper"). The "White Paper" and the "Digital Intelligence Materia Medica" model were released on May 9.

In recent years, the mainland has formulated a series of policies to promote the digital and intelligent development of the traditional Chinese medicine industry. However, in the current era of digitalization and intelligence, the development of big data and artificial intelligence technology has brought unprecedented opportunities and challenges to various fields. The white paper believes that the modern Chinese medicine industry cannot be limited to the traditional production and research mode, and the promotion of digital intelligence is the general trend. However, the development of digital intelligence in modern Chinese medicine is not a smooth road. The development of digital intelligence in modern Chinese medicine is inseparable from the accumulation and analysis of a large amount of data, but how to ensure data security and patient privacy is a serious challenge. With the rapid progress of TCM production technology and the changes in the market demand of TCM, it is necessary to continuously optimize and improve the existing algorithms to improve the accuracy and efficiency of the algorithms. The development of digital intelligence in modern Chinese medicine requires a large amount of computing resources, and how to effectively integrate and utilize computing resources is a challenge. There are still many problems in the connection between digital intelligence Chinese medicine and new drug R&D paths, such as the difficulty of data acquisition and uneven quality; Difficulty in algorithm selection and optimization; Lack of quality control standards, etc.

In order to solve the problems and challenges faced by the development of digital intelligence in modern Chinese medicine, the white paper believes that it is necessary to strengthen the standardization and normalization of traditional Chinese medicine data, establish a traditional Chinese medicine data sharing platform, and ensure data security while improving data quality control. At the same time, it is necessary to continuously develop and optimize the algorithm to improve the accuracy and effectiveness of the algorithm. Integrate resources and enhance computing power to provide a guarantee for the development of digital and intelligent Chinese medicine. In addition, it is necessary to clarify the connection between the pathway between digital intelligence and new drug research and development to improve the efficiency and quality of Chinese medicine research and development.