Develop new productivity and artificial intelligence to empower the future


The Central Economic Work Conference proposed to accelerate the development of artificial intelligence. Artificial intelligence is a technology that simulates, extends and expands human intelligence. Simply put, it allows machines to learn human wisdom and respond intelligently like humans. Artificial intelligence is a strategic technology that leads the future, the core driving force of a new round of scientific and technological revolution and industrial transformation, and is considered to be the main position for developing new quality productivity. How is China's artificial intelligence technology and industry developing?

       The Central Economic Work Conference proposed to accelerate the development of artificial intelligence. Artificial intelligence is a technology that simulates, extends and expands human intelligence. Simply put, it allows machines to learn human wisdom and respond intelligently like humans. Artificial intelligence is a strategic technology that leads the future, the core driving force of a new round of scientific and technological revolution and industrial transformation, and is considered to be the main position for developing new quality productivity. How is China's artificial intelligence technology and industry developing?


       In the past 2023, generative artificial intelligence was undoubtedly the most dazzling "technological star". Domestic artificial intelligence large models continued to iterate and upgrade, and their ability to interact with natural language and generate multi-scenario content was close to that of humans.

       By the end of 2023, my country has released more than 200 AI big models, of which more than 20 big model products have been approved to provide services to the public. In fact, most of the big models available to the public are general-purpose base big models. Their abilities are not only to write poems, paint, solve math problems, and make PPTs, but their greater ability is to enable intelligent industrial production and accelerate the formation of new quality productivity.

       Wang Haifeng, Baidu's chief technology officer and director of the National Engineering Research Center for Deep Learning Technology and Applications: "We have trained a large model, which is equivalent to it having strong general knowledge. When fine-tuning for a certain application scenario, perhaps dozens or hundreds of data are enough. Once the basic model is constructed, it can be quickly promoted to various industries at low cost."

       The reason why the big model is as smart as a human being and can use logical reasoning to generate text, pictures, sounds, videos, codes and other content is because it has been trained in deep learning on a massive amount of information knowledge. The variable parameters of this information are very large, reaching tens of billions or trillions.

       Wei Kai, executive deputy director of the Artificial Intelligence Research Center of the China Academy of Information and Communications Technology, said: "When the model exceeds a certain scale, it will have the ability to emerge with intelligence. It can learn things that it has never been taught. It is just like piling up computing power and data. When the scale goes up, the performance will continue to improve."

       iFlytek is one of the leading companies in my country's artificial intelligence field. In May 2023, iFlytek released a base-type general-purpose large model - the iFlytek Spark cognitive large model. Immediately afterwards, they took the industrial scene as the connection point and the general capabilities of iFlytek Spark as the core technology base, and quickly launched the Antelope Industrial Large Model, which has five core capabilities including industrial text generation, industrial knowledge Q&A, industrial understanding computing, industrial code generation, and industrial multimodality, and can continuously empower the entire manufacturing process. Based on the Antelope Industrial Large Model, they also focused on creating the "Digital Craftsman" artificial intelligence application product.

       Xu Jiajia, President of Antelope Industrial Internet Co., Ltd.: "What does a 'digital craftsman' do? There is often an old master in a factory who knows everything we ask him because he has accumulated a lot of industrial knowledge and experience in this position. For example, the digital surveyor who cooperates with China Classification Society is a knowledge-intensive position with a very long training cycle. We can train a lot of ship inspection knowledge into it and provide guidance for ship inspection through large models, which can assist the surveyor in better performing ship inspection work."

       Big models are a new engine driving the intelligentization of industries, and manufacturing is the most important area for artificial intelligence to accelerate the formation of new quality productivity. The first "Lighthouse Factory" in the global special steel industry, located in Jiangyin City, Jiangsu Province, is known as the world's most advanced factory and represents the top intelligent manufacturing capabilities in the industrial field. One of the important reasons why this company was named a "Lighthouse Factory" last year is that they used a large number of artificial intelligence technologies such as big models in their digital and intelligent transformation, which greatly improved the level of intelligence and production efficiency.

       Wang Xiaomin, Intelligent Supervisor of Xingcheng Special Steel Thick Plate Branch of CITIC Pacific Special Steel Group: "There are thousands of high-precision testing instruments all over the site, which are equivalent to our five senses. They can obtain the status of the production line in real time. On this basis, we have established an intelligent model based on industrial mechanisms and AI. Through the continuous self-learning of the model, we can obtain the optimal guidance strategy to guide production. At present, our labor productivity has increased by 29.8%, and the quality loss cost has decreased by 46.7%."

       This "lighthouse factory" can produce 6.9 million tons of special steel every year. By combining artificial intelligence technology and high-precision sensing instruments, they have solved the contradiction between the special steel industry's multi-variety, small batch and large-scale production, and can achieve flexible customization of products. At present, the number of customized orders for enterprises has increased by 35.3%.

       Wang Xiaomin: "Companies use industrial Internet technology to collect field data in real time and aggregate it to cloud platforms, where they deploy artificial intelligence models. These massive amounts of data can be used to train more powerful and accurate models to cope with complex and changing production tasks."

       As a strategic technology and general technology leading the future, artificial intelligence is considered an important engine for the formation of new quality productivity. Since the 18th National Congress of the Communist Party of China, my country has seized the opportunity to deploy artificial intelligence from the national strategic level, and the policy has been timely and precise, achieving remarkable results. Statistics show that my country ranks first in the world in the number of papers published in the field of artificial intelligence, the scale of the core industry of artificial intelligence has reached 500 billion yuan, and the number of enterprises has exceeded 4,400. Industrial quality inspection, knowledge management, autonomous driving, voice interaction... Artificial intelligence applications are evolving in depth. In the new technological track of artificial intelligence, my country is already in the world's first echelon.

       Wei Kai: "The reason why artificial intelligence has made such a leap forward is mainly due to massive data, the ability to learn from data and gain intelligence. This means that the key to future success is whoever has abundant data will be able to master artificial intelligence. It is precisely our country that has such an advantage. We have a massive population and a large number of Internet users, which will accumulate extensive and diverse data resources on the consumer side. In addition, we have the most complete manufacturing system in the world, and have accumulated a large amount of high-value data on the production side."

       Data, computing power, and algorithms are considered to be the important technical foundations for the development of artificial intelligence. The rapid development of my country's artificial intelligence industry also benefits from my country's advanced layout of digital information infrastructure such as computing power. In 2022, the total scale of data center racks in use nationwide exceeded 6.5 million standard racks. More than 30 cities are building or proposing to build intelligent computing centers, and the total scale of computing power has ranked second in the world. 

       Wei Kai: "Data is like the raw materials in artificial intelligence, and algorithms are equivalent to the formula and process flow for processing raw materials. Algorithms are run by computing power. For example, large models today need to feed massive amounts of data to large models, which requires a large computer cluster to process these huge amounts of data. Often, tens of thousands or even hundreds of thousands of machines may be needed to train such a model together."

       In the application scenarios of artificial intelligence, virtual digital humans are an important application direction. In recent years, with the rapid development of the digital economy, various types of digital virtual anchors, digital commentators, digital human intelligent customer service, and AI avatars have emerged. Now, with the rapid penetration of large model technology, digital human products have also ushered in a leap in intelligence.

       Some experts predict that in the future, not only digital humans will assist humans in their work, but also a large number of robots powered by artificial intelligence technology. The future production model may be a division of labor and collaboration among three types of parallel employees: "natural humans + robots + digital humans".

       The maintenance workshop of the Nanning South Depot of the Nanning National Railway Bureau is busy during the Spring Festival travel rush. Deputy foreman Lin Xing's daily workload is to inspect the pillow springs of more than 30 freight car bogies, with a total of 3,000 to 4,000. This was an impossible task in the past. In the past, it took three workers a day to manually inspect the pillow springs of 30 cars. But now Lin Xing is confident because he has a new assistant - the pillow spring intelligent detection and selection robot. This new assistant uses a number of artificial intelligence technologies, which can accurately calculate and analyze the diameter, height and other information of the pillow springs, and quickly select unqualified pillow springs.

       Robotics is an important field for the deep integration of artificial intelligence technology and the real economy of manufacturing. In recent years, my country's robotics industry has developed rapidly, with the installed capacity of industrial robots ranking first in the world and the density of robots in the manufacturing industry reaching 392 per 10,000 workers. With the development and maturity of many technologies such as large models, humanoid robots, which represent the direction of industrial robot upgrades, have become a hot field for building future industries and developing new quality productivity. Humanoid robots are expected to become another disruptive product after computers, smart phones, and new energy vehicles.

       Wei Kai: "Large models will accelerate the evolution of the humanoid robot's brain, making it smarter and able to cope with very complex scenarios. Once these humanoid robots can be mass-produced, they will open up a new world for our production and life. If the productivity leap of the previous generation relied on energy and information, then the productivity leap of the next generation will rely on intelligence. Seizing the opportunity of this wave of general artificial intelligence means that we can seize the future."

       Now, our lives are inseparable from artificial intelligence. When you search for information on the Internet, when you go out and use a map to navigate, when you enter a community and use facial recognition to open the door, artificial intelligence technology is used. Artificial intelligence, like the invention of steam engines, electricity and network technology, has become the core driving force of the fourth industrial revolution of mankind, and will eventually affect all aspects of human production and life. The future is here. Only by vigorously developing artificial intelligence technology, promoting the deep integration of artificial intelligence and the real economy, and accelerating the formation of new quality productivity can we seize the leading position in the new round of international scientific and technological competition and provide more powerful support for promoting Chinese-style modernization.