Recently, the State Council issued the "Opinions on Deepening the Implementation of the" Artificial Intelligence+"Action" (hereinafter referred to as the "Opinions"), which clarifies the overall requirements for promoting the "Artificial Intelligence+" system, systematically lays out the six key actions and eight basic supporting capabilities of "Artificial Intelligence+", and points out the implementation path for the deep integration of artificial intelligence with the economy and society in China. The release of the "Opinions" marks a new stage of systematic promotion and deep integration in the development of artificial intelligence in China, which helps to comprehensively transform China's advantages in industrial scale, data resources, application scenarios, etc. into national intelligence advantages. In the implementation process, it is necessary to adhere to systematic innovation as the core, and through the triple innovation coordination of technology, industry, and mechanism, focus on breaking through cutting-edge fields such as industrial models, scientific models, low altitude economy, and embodied intelligence, and build a benign innovation ecology of "technology industry society" development. The Opinion on Strengthening the Innovation of Scientific Models and Accelerating the Transformation of Scientific Research Paradigms places "artificial intelligence+science and technology" as the top priority, reflecting the country's profound grasp of the development law of artificial intelligence: only by consolidating the scientific foundation can we grasp the initiative in technological breakthroughs, industrial upgrading, and global competition. Scientific big models are an important manifestation of the intersection and integration of artificial intelligence and basic science. It has demonstrated disruptive potential in fields such as protein structure prediction, astrophysical simulation, and new material discovery. By constructing a large-scale model architecture for scientific reasoning and discovery, the entire process of scientific research from hypothesis formulation to experimental verification can be significantly accelerated, and even the development paradigm of certain fundamental disciplines can be reconstructed. It is recommended to prioritize supporting the research and development of large-scale models and the construction of open platforms for major scientific issues, promoting the joint construction of scientific intelligent infrastructure by high-level research institutions, universities, and enterprises, especially focusing on the integration of key capabilities such as cross modal reasoning, and creating an intelligent foundation to support cutting-edge scientific discoveries. Promoting innovation in the application of industrial big models to empower the transformation and upgrading of thousands of industries. Currently, artificial intelligence technology is rapidly evolving from general big models to industrial big models. This transformation not only expands the boundaries of technology itself, but also profoundly reshapes the research and development models and paths of thousands of industries. Through pre training with massive industry data and deep embedding of domain knowledge, industrial big models fully leverage their powerful natural language understanding and generation capabilities, significantly enhancing enterprises' cognition and understanding of complex business scenarios. The deep integration of industry knowledge and scenario data is becoming the core engine driving the intelligent transformation of industries. In the research and development process, industrial big models can accelerate technological insights and knowledge discovery. Through intelligent analysis of global patents, papers, and technology reports, they help enterprises quickly grasp technological development trends and optimize research and development routes. In the production process, applications such as intelligent quality inspection, process optimization, and equipment predictive maintenance based on large models are significantly improving production efficiency and product quality. The Opinion clearly proposes to promote the integrated and coordinated development of technology research and development, engineering implementation, and product landing driven by artificial intelligence, and the industrial model is the key lever to achieve this goal. In cutting-edge fields such as biomanufacturing, quantum technology, 6G, and new materials, industrial models use intelligent algorithms to significantly reduce research and development cycles and improve the efficiency of innovation transformation. For example, in the development of new materials, large models can predict material properties and reduce experimental trial and error costs; In the field of equipment manufacturing, large models support integrated design and simulation, accelerating product iteration and upgrading. To promote the deep application of industrial models, it is recommended to focus on strengthening three aspects of work. One is to build a large-scale model system for enhancing industry knowledge and develop a vertical domain oriented model optimization toolchain; The second is to promote the deep integration of research, production, and application, and establish a collaborative innovation mechanism guided by enterprise needs and supported by large-scale models; The third is to improve the evaluation standards for industrial large-scale models, ensuring that the model performance matches industry needs. The layout of low altitude large-scale models and embodied large-scale models to cultivate new growth in the era of intelligence, as an important breakthrough in the "AI+" action, represents not only the forefront direction of technological application, but also a key area for cultivating new quality productivity and shaping new development momentum, which will have a profound impact on China's economic transformation and upgrading. The breakthroughs in these two types of vertical domain models will drive the evolution of artificial intelligence from cognitive intelligence to action intelligence, providing key technological support for cultivating new quality productivity. Low altitude large-scale models are the core infrastructure for achieving intelligent upgrading of the low altitude economy, and multimodal large-scale models with spatial cognitive capabilities are a potential key research and development direction. It integrates knowledge from geographic information, meteorological data, flight rules, and other fields to provide environmental perception, path planning, and risk warning capabilities for low altitude aircraft. The specific implementation path includes: building a pre trained large model in the low altitude domain, injecting professional knowledge such as airspace management and flight control; Develop a group decision-making system based on large models to achieve autonomous collaboration and conflict resolution of drone clusters; Establish a low altitude large-scale model testing and verification platform, and implement algorithm iterative optimization through digital twin technology. At the same time, safety standards for low altitude large-scale models should be established to ensure the interpretability and reliability of the decision-making process, providing reliable artificial intelligence support for the development of low altitude economy. The embodied intelligent big model is a key bridge that connects virtual intelligence with the physical world. Efforts should be made to develop embodied cognitive models that integrate multimodal information such as vision, language, and action to achieve deep understanding and task execution of physical scenes by intelligent agents. The key breakthrough directions include: developing tailored decision-making models to support the decomposition and planning of complex tasks; Develop a large-scale human-machine collaboration model to achieve precise conversion from natural language instructions to physical actions; Build a personalized training simulation platform to enhance the model's generalization ability through large-scale interactive data. In terms of industrial implementation, we should take the lead in carrying out application demonstrations in industrial testing, home services, medical rehabilitation and other scenarios, and cultivate typical application paradigms of embodied intelligent large models. Building a collaborative innovation system to promote the strengthening of basic research and intelligent base construction in the "Artificial Intelligence+" action. Focusing on scientific models, general artificial intelligence, and embodied intelligence, we will focus on promoting the research and development of large-scale models oriented towards fundamental theories and key architectures. Support the construction of a national level AI platform and open source community, promote high-level research institutions and enterprises to jointly build a collaborative R&D environment based on large models, promote the integration and sharing of models, algorithms, and development tools, and provide a solid intelligent foundation for the "AI+" action. Promote interdisciplinary and cross-border integration. We strongly support the deep integration of large-scale models with life sciences, materials science, social sciences, and other fields to create large-scale models for major scientific problems and engineering challenges. Promote big models to become interdisciplinary "connectors" and "accelerators", support the construction of collaborative research platforms based on big models, encourage the formation of interdisciplinary research teams combining artificial intelligence and disciplines, and generate significant interdisciplinary innovation achievements. Improve the innovation ecosystem and governance system. Accelerate the establishment of standards, ethics, and governance systems covering the entire lifecycle of large models, with a focus on promoting the compliant application of large models in sensitive fields such as low altitude economy, autonomous driving, and medical AI. Support the regulatory construction of big model data compliance, evaluation benchmarks, security and controllability, and build a governance framework with Chinese characteristics that adapts to the innovation needs of big models, creating an open, secure and trustworthy innovation environment for the "AI+" initiative. The issuance of the 'Opinions' will undoubtedly accelerate the vigorous development of' artificial intelligence+'in China. As frontline workers in the field of artificial intelligence research, we will continue to focus on key directions such as scientific big models, industrial big models, low altitude economy, embodied intelligence, etc., promote the construction of a systematic innovation layout, and help China's "AI+" action deepen and become practical. (New Society)