Recently, the Ministry of Industry and Information Technology and eight other departments jointly issued the "Implementation Opinions on the Special Action of 'Artificial Intelligence+Manufacturing'", which systematically deploys artificial intelligence to empower the manufacturing industry based on technological foundations, application scenarios, and industrial ecology. The deep integration of artificial intelligence and manufacturing industry can not only improve production efficiency and reduce manufacturing costs, but also promote the full process transformation of research and development design, production organization, quality control, supply chain collaboration, etc., gradually shifting the manufacturing industry from factor driven and experience driven to data-driven and intelligent decision driven. Accelerating the promotion of "artificial intelligence+manufacturing" is not only an essential part of cultivating new quality productivity, but also a necessary path to enhance the resilience of industrial and supply chains and improve the core competitiveness of the manufacturing industry. China's manufacturing industry has a complete range of categories, a complete chain, and rich scenarios, which have natural advantages in promoting the implementation and effectiveness of artificial intelligence. At the same time, it should be noted that the intelligent transformation of the manufacturing industry is still in the stage of climbing hills and overcoming obstacles. How to connect data and optimize processes after installing devices and systems, so that intelligence does not remain in the stage of local automation? How to develop solutions, form a stable talent pool, and achieve economies of scale with platforms and projects? There are still many issues that need our attention and resolution. The more we are at such a critical stage, the more we need to have a clear direction and a solid path, adhere to a systematic approach, and continue to work hard to grasp the fundamental and guiding tasks. To promote the deepening and implementation of "artificial intelligence+manufacturing", it is necessary to not only lay a solid foundation of independent and controllable technology and data, so that artificial intelligence can operate stably in industrial scenarios, but also to expand intelligent applications from single point exploration to overall promotion through industrial chain collaboration. At the same time, it is necessary to improve the integration of industry and education and talent training mechanisms, providing continuous support for the transformation and upgrading of the manufacturing industry. Promoting 'artificial intelligence+manufacturing' relies on technology as the foundation and data as the key. The manufacturing industry has extremely high requirements for stability and safety, and any fluctuations may bring quality and safety risks. Therefore, when artificial intelligence enters the manufacturing industry, it not only needs to be usable, but also stable and long-lasting. This first depends on whether the technical foundation is solid and whether the data system is sound. On the one hand, we must firmly grasp the key core technologies in our own hands. The intelligence of manufacturing industry cannot be separated from computing power and industrial software support, especially in the fields of high-end chips, core software, etc. Only by breaking through the technological blockade of core industrial software, realizing independent research and iteration from basic software to industry application software, can the "brain" of manufacturing industry intelligence truly operate independently. We must continue to promote independent innovation, enhance the deployment and operational capabilities of artificial intelligence in industrial sites, and ensure the long-term stable operation of the system. On the other hand, we need to build the basic resource of industrial data well. The effectiveness of artificial intelligence models largely depends on data quality. If enterprise data is still scattered across different devices and systems, with inconsistent standards and uneven quality, it can lead to problems such as "too much data is collected, too little is used". The next step should be to promote both the technological foundation and the data system simultaneously, and to form effective data resources around research and development, production, quality, and operation and maintenance, providing long-term support for the intelligence of the manufacturing industry. Promote the integration of artificial intelligence and manufacturing, with a focus on collaboration and addressing challenges through connectivity. The intelligent transformation of individual processes and production lines is certainly important, but what truly determines the competitiveness of the manufacturing industry is the full chain linkage of design, research and development, production, supply chain, and service. To move from local efficiency improvement to system efficiency improvement, artificial intelligence needs to break down the data barriers within the enterprise and connect the collaboration chain between upstream and downstream. In practice, some regions have accelerated the implementation of new technologies through platform construction and scenario integration; Some enterprises improve production line switching and promotion efficiency through demonstration simulation and cloud replication. Next, we should take industrial chain collaboration as an important lever to promote intelligence, leverage the leading role of leading enterprises, and promote mature and effective solutions to more small and medium-sized enterprises. Focus on key industries, summarize a batch of typical application scenarios, summarize successful experiences, continuously optimize and promote in practice, and avoid intelligent transformation, fragmentation, and islanding. The success or failure of promoting "artificial intelligence+manufacturing" lies in talent and mechanisms. The intelligentization of manufacturing industry is a long-term project that requires a composite talent team that understands both process equipment and data algorithms. Technicians understand the field and frontline personnel are proficient in systems, so that applications can be implemented and continuously iterated and upgraded. The most difficult part of industrial sites is not writing algorithms, but embedding algorithms into processes, models into management, and data into production. Ultimately, it tests the comprehensive abilities of enterprise employees in technology, management, and engineering. Next, the integration of industry and education should be placed in a more prominent position, promoting the joint construction of industrial colleges, laboratories, and training bases between universities and enterprises, and allowing real industrial problems to enter the curriculum and scientific research; Improve the on-the-job training system for enterprises and promote the transformation of industrial workers towards operation and maintenance, debugging, data processing, and process optimization; Optimize the talent evaluation mechanism, encourage interdisciplinary cooperation and the cultivation of job composite abilities, cultivate more "digital engineers who understand the industry" and "on-site engineers who understand data", and promote enterprises to shift from one-time project construction to long-term capability accumulation. Overall, adhering to the principle of seeking progress while maintaining stability, promoting research through application, and using points to lead the way can ensure the stable use of artificial intelligence in the workshop, transform single point exploration into universal capabilities, and form an internal driving force for industrial transformation and upgrading. In this way, China's manufacturing industry will continue to enhance its resilience and competitiveness, providing more solid support for promoting new industrialization and achieving high-quality development. (New Society)