Data sources: CAICT, IDC
With the digital infrastructure construction continues to improve in recent years, the commercialization of applications accelerates to landing and artificial intelligence industry quickly develops.
Along with industrial quality inspection, parts counting, autonomous driving, and voice interaction, artificial intelligence in product development, service upgrading, business model innovation and other aspects of the enterprise brings tangible results. A variety of artificial intelligence applications enter into production life and service thousands of households.
The new generation of artificial intelligence is to promote the leapfrog development of science and technology, industrial optimization and upgrading in productivity. At present, the field of artificial intelligence in China presents a two-wheel drive of technological innovation and industrialization applications of the development characteristics.
With the joint promotion of both supply and demand, the results of technological innovation began to move from laboratory research to industrial practice on a large scale. The industrialization of artificial intelligence accelerated. According to the latest data released by the China Academy of Information and Communications Technology, China's AI core industry scale reached 508 billion yuan in 2022, an increase of 18% year-on-year.
Accelerate application breadth and expansion, scene innovation becomes the new path
In recent years, the domestic Jian tea cup industry rapidly develops, but also exists counterfeit imitation, substandard and other chaos. It is difficult to solve the problem through traditional anti-counterfeiting means.
Later, a Jian tea cup production enterprise located in Jiangyang District, Nanping City, Fujian Province, found Beijing Megvii Technology Co.
This is not an easy task. On the one hand, it is extremely difficult to distinguish the difference between different Jian tea cup marigolds. The glaze of marigolds also produces strong reflections, all of which can pose challenges for artificial intelligence; On the other hand, such projects often require a lot of time for experimental design.
In the face of these problems, Megvii team proposed the concept of "algorithm mass production" and created an artificial intelligence traceability system to achieve the identification of marble patterns. The team also developed a hardware device with built-in controllable light source to overcome the problem of glaze material reflection.
Erjin Zhou, head of algorithm mass production of Megvii Research Institute said, "We have thus realized a picture, a code and a combination of picture, which not only effectively combats imitation and forgery, but also promotes the standardized development of the industry and the upgrade of digital intelligence."
"AI technology is continuing to evolve along the pursuit of higher accuracy, challenging more complex tasks and expanding the boundaries of capabilities. Scene innovation has become a new path for AI technology upgrade and industry growth." Zhao Zhiying, director of the Research Center for the Development of New Generation Artificial Intelligence of the Ministry of Science and Technology and secretary of the party committee and director of the China Institute of Scientific and Technological Information, believes that the breadth and depth of AI applications are expanding and are penetrating into more industries and more core business areas, driving the increasing intelligence of manufacturing, logistics, healthcare and other aspects. The importance of intelligent scenarios for economic and social development is gradually coming to the fore.
New Technology gradually emerges, industry path becomes clear
With the continuous emergence of new technologies, the path of AI industrialization is also gradually clear. In recent years, domestic and foreign manufacturers have increased the input and research and development of huge models, so that the industrial implementation of artificial intelligence has found a new direction.
New drug development benefits from this. Huawei Cloud Pangu Drug Molecular Big Model, a big model jointly trained by Huawei Cloud and the Shanghai Institute of Pharmaceutical Sciences of the Chinese Academy of Sciences, can realize artificial intelligence-assisted drug design for the whole process of small molecule drugs.
Professor Bing Liu's team at the First Affiliated Hospital of Xi'an Jiaotong University has successfully developed a new "super antibacterial drug" using the Big Model, which is expected to be the world's first new target and class of antibiotics in nearly 40 years. With the help of the Big Model, the development cycle of the lead drug has been shortened from several years to one month, and the development cost has been reduced by 70%.
According to Rui Yong, chief technology officer of Lenovo Group, the characteristics of the big model can be summarized as “one big and three multi”: “One big” refers to the large scale of parameters, which is a super-large AI model with hundreds of billions of parameters; “Three multi” refers to the use of multi-source, multi-modal, multi-task Internet mass data for training.
"On the basis of cloud services, the landing of big models is also becoming more and more common. In the past, we often emphasized small models for a certain type of application, such as face recognition and speech recognition. Now with the advent of big models, users can perform simple processing on pre-trained models to meet their needs. This can actually help users to better apply AI technology." said Zhen Gang Zhou, vice president of International Data Corporation (IDC) China.
The intelligent customer service robot "Yuan Xiao Service" built by Wave Information based on the big model is a typical application of the big model. Chen Bin, service director of Wave Information said, "Our aim is to combine the big model with the industry model that fits the needs of the scene."
With its powerful learning capability, "Source Dawn Service" is able to learn the knowledge base autonomously. At present, it has been able to cover 92% of end-users' consulting problems. For some common technical problems, such as system installation and component abnormalities, the solution rate reaches 80%.
"The large model is not a show-off, rather it is more derived from the realistic demand of AI industry development." Liu Jun, vice president of Wave Information, pointed out that AI application scenarios have transitioned from fragmentation to deep integration. In terms of algorithm models, huge models will become the basis for scaled-up AI algorithm innovation.
Rapid growth in the scale of core industries, forming a good foundation for industrial development
Artificial intelligence is a strategic technology that leads the new round of technological revolution and industrial change. In recent years, China has introduced a number of policies to encourage the development and innovation of the artificial intelligence industry. Industrial policies such as the "Development Plan for New Generation Artificial Intelligence", "Notice on Supporting the Construction of New Generation Artificial Intelligence Demonstration Application Scenarios" and "Three-Year Action Plan for the Development of New Data Center (2021-2023)" have provided guarantees for the development of AI industry. China has also relied on leading enterprises to build more than 10 national open innovation platforms for new generation AI.
In recent years, China has formed a good foundation for the development of artificial intelligence industry. In addition to the rapid growth in the scale of the core AI industry, the number of AI enterprises in China exceeds 4,200. Key core technologies such as smart chips and open source frameworks have made important breakthroughs, and the innovation capacity of iconic products such as smart sensors and smart network-connected vehicles continues to improve. In 2022, the total scale of data center racks in use nationwide exceeds 6.5 million standard racks. More than 30 cities are currently building or proposing to build intelligent computing centers.
Experts say that the key to further promoting the development of AI industry is transforming innovation elements into material or knowledge resources. Forming scale and scope effects lies in the construction of collaborative innovation platform. The construction of AI collaborative innovation platform should focus on the current demand for landing applications. It is important to realize business aggregation, resource aggregation and strategic aggregation of AI ecological partners. We need to form an industrial ecology with full cooperation from upstream and downstream of the industry chain by strengthening the supply of AI arithmetic output, service capacity optimization and talent cultivation in the platform.
Source: People's Daily