Artificial intelligence is an important engine for developing new quality productivity, providing significant historical opportunities for creating new advantages in the manufacturing industry in the era of intelligence. This year's government work report proposes to create a new form of intelligent economy, providing clear direction and strong support for the intelligent transformation of industries. Currently, compared to the comprehensive outbreak on the consumer side, industrial intelligence is still in its infancy. Physical enterprises with industrial scenarios do not understand how to integrate AI technology, while technology enterprises with AI technology do not understand industrial scenarios; Lack of benchmark guidance for "artificial intelligence+manufacturing" in industrial intelligence; Industry level intelligent industry ecological infrastructure is weak; There is a shortage of compound talents who understand industries and AI... To solve these challenges, we need to start from the following aspects. Systematic research defines the top-level architecture of "artificial intelligence+manufacturing". In the era of AI, the underlying logic of manufacturing may undergo fundamental changes, forming a new way of "industry brain empowerment+intelligent agent collaborative promotion", which requires the reconstruction of technology research and development paradigms, supply chain management models, and intelligent manufacturing systems. Starting from key industries, we should mobilize manufacturing enterprises and AI technology enterprises to jointly create industry level business and technology architectures, define a new model of the next generation industry centered on industry brains and intelligent agent platforms, and integrate multiple AI technology construction paths to form top-level leadership. Accelerate the cultivation of benchmarks and standards for "artificial intelligence+manufacturing" and lead the upgrading of industrial intelligence. Create industry benchmark projects, accelerate the implementation and application of intelligent agents in industrial scenarios, and form replicable and promotable practical models; On the basis of benchmark practice, establish evaluation criteria to provide path guidance for the upgrading of industrial intelligence. Build an industry level 'intelligent agent brain' and strengthen the new infrastructure of the intelligent industry ecosystem. Promote joint research and development among artificial intelligence enterprises, leading companies, and research institutions, focusing on manufacturing data platforms, vertical models, and industry shared knowledge bases, to create an "intelligent agent brain"; Encourage leading enterprises and artificial intelligence companies to jointly develop industry data collection, governance, and sharing standards, in order to overcome the bottleneck of scattered data and diverse standards; Encourage leading enterprises to open up industry data assets and explore cross industry and cross regional data sharing mechanisms. Establish a comprehensive talent training and certification system, with a focus on cultivating industry AI architects who understand industries and AI. Develop standards for identifying the abilities of composite talents in the era of artificial intelligence, clarify the training direction and evaluation criteria for industrial AI architects; Promote two-way talent flow, establish an integrated training mechanism for industry, academia, research and application, encourage linkage between technology enterprises and industrial enterprises, carry out full chain talent training, and accelerate the large-scale supply of composite talents. (New Society)