CITIC SEC: Multi-modal AI capabilities upgrade, Industrial AI leading the Fourth Industrial Revolution.

CITIC SEC released a research report stating that the industrial sector is a pillar industry in China, and industrial digitalization is the key to achieving new industrialization. Currently, China's industrial digitalization development still faces issues of insufficient and uneven integration, but multimodal AI is expected to provide new solutions, driving the transformation of industrial intelligence. On the demand side, the industrial AI market is showing strong growth momentum, with the China Academy of Information and Communications predicting a global market size of 895 billion dollars by 2032. On the supply side, there are a variety of players in AI+ industry, including major model manufacturers, traditional software companies, and large manufacturing companies, with a suggestion to focus on enterprises that are the first to achieve seamless integration of software and hardware. The main points of CITIC SEC are as follows: China's level of industrial digitalization is lower than the global average, facing issues of insufficient and uneven integration. Industry is a pillar industry in China. According to the National Bureau of Statistics, in 2023, China's industrial value added was 39.9 trillion yuan, accounting for 38% of that year's GDP, providing stable momentum for economic growth. Industrial digitalization is the key to achieving new industrialization, which is expected to improve industrial production efficiency and China's independent controllable capabilities. The national government has issued several policies to support this, providing a broad development space for the industry. However, due to the incomplete systematization of China's industrial digitalization development, the penetration rate is far below the global average (according to the Ministry of Industry and Information Technology, China's manufacturing digital penetration rate was 24.7% in 2022 vs the global average of 33%), and there are significant differences in the digitalization process among different industries. For example, high value-added industries such as semiconductors, automobiles, and aerospace lead the way in development, while industries like light industry and textiles are slow in transformation. Therefore, China still faces certain development bottlenecks and pain points in the industrial digitalization transformation. Multimodal AI large models are expected to bring a new paradigm of "basic models + various applications" to the industrial field, thereby increasing the popularity of AI applications in the industrial sector and empowering the transformation of industrial intelligence. Industry is an important landing scenario for multimodal AI, driving the industrial sector into the intelligent stage. In terms of feasibility, data-driven is a common feature of industrial digitalization and AI. The industrial sector has already established a foundation for industrial digitalization development with abundant data resources, providing possibilities for the rapid landing of AI+ industry. In terms of scenarios, industrial scenarios can be divided into six categories: research and development design, production manufacturing, warehousing logistics, management, product services, and industrial safety. AI+ industry can be applied in energy and power, automobile manufacturing, aerospace, and other industries. According to Tencent Research Institute, large models are more suitable for research and development design and management, and their capabilities in production and manufacturing still need to be improved. With the improvement of accuracy, industry know-how, data accumulation, and application capabilities of large models, AI+ industry is expected to continue to expand to multiple links of the industrial chain. In terms of utility, the empowerments of AI+ industry lie in cost reduction, efficiency improvement, safety and stability, and lowering thresholds. For example, the development of multimodal large models can bring machine vision, NLP, and speech recognition applications to industrial design software, improving production efficiency and reliability, lowering professional thresholds, and helping China achieve a leap in industrial intelligence. The development of multimodal AI+ industry is accelerating, with the singularity of a new industrial revolution approaching. From a technical perspective, multimodal AI is currently in its early stages. With the accelerated iteration of large model capabilities, such as the improvement of long text capabilities, enhancement of video generation capabilities, and improvement of speech interaction capabilities, large model AI is expected to further expand its applications in the industrial sector. From a business perspective, according to the Financial Magazine Business Review, the prices of global manufacturers' main model inferences have significantly decreased. In terms of applications, compared to general AI, the lag period for industrial application AI has been shortened from 20 years to less than 5 years, making the future development of AI+ industry worth looking forward to. From a market perspective, on the demand side, the industrial AI market is showing strong growth momentum, with the China Academy of Information and Communications predicting a global market size of 895 billion dollars by 2032. On the supply side, there are a variety of players in AI+ industry, including major model manufacturers, traditional software companies, and large manufacturing companies, each with its own advantages, suggesting to focus on enterprises that are the first to achieve seamless integration of software and hardware. Risk factors: Risks of artificial intelligence and multimodal large model technology development falling short of expectations; Risks of policy support not meeting expectations; Risks of intensified industry competition; Risks of macroeconomic fluctuations affecting corporate R&D spending. Investment strategy: Industry is a pillar industry in China, with industrial digitalization being the key to achieving new industrialization. Currently, China's industrial digitalization development still faces issues of insufficient and uneven integration, but multimodal AI is expected to provide new solutions, driving the industry into a new phase of intelligent development. On the demand side, the industrial AI market is showing strong growth momentum, with the China Academy of Information and Communications predicting a global market size of 895 billion dollars by 2032. On the supply side, there are a variety of players in AI+ industry, including major model manufacturers, traditional software companies, and large manufacturing companies, each with its own advantages, suggesting to focus on enterprises that are the first to achieve seamless integration of software and hardware.