Cultivating new quality productivity.
The theme of this year's 517 World Telecommunication and Information Society Day is "Digital Innovation for Sustainable Development", highlighting the key role of innovation in digital technology and applications for sustainable economic and social development. With the accelerated evolution of artificial intelligence technology, AI large models have become a new highland for global scientific and technological competition, a new track for future industries, and a new engine for economic development.
In recent years, the mainland has attached great importance to the development of artificial intelligence, elevated it to a national strategy, and introduced a series of supportive policies and plans, creating a good environment for the development of the AI large model industry, and the large model has become the biggest hot spot in the field of artificial intelligence. How to effectively integrate large model technology into the practical application of all walks of life and help productivity innovation and industrial upgrading has become a core issue that the industry is increasingly concerned about, and has also promoted the development of large models in the industry.
Build a new quality of productivity together
"Vigorously promote the construction of a modern industrial system and accelerate the development of new quality productive forces." It is listed as the top ten tasks in the 2024 government work report.
In fact, the essence of "new quality productivity" is to use scientific and technological innovation to drive productivity to a new quality state to achieve "new quality improvement". In this process, the new generation of artificial intelligence, represented by large models, is not only the core driving force of the digital economy, but also an important technical foundation for "new quality productivity". It is foreseeable that as the large model continues to "go deeper and more practical", it will not only promote a new round of scientific and technological revolution and industrial transformation, but also provide key support for the digital and intelligent transformation of thousands of industries.
Data shows that generative AI alone could add between $2.6 trillion and $4.4 trillion annually to the global economy by 2040. At that time, from individuals to enterprises, from life to production, from the edge to the core, all scenarios, processes, and businesses will be reshaped.
Industry experts pointed out that with the continuous upgrading of domestic AI large models, 2024 is expected to be the first year of the full commercial implementation of domestic AI large models, and the advanced AI large model capabilities will gradually be transformed into product applications in actual scenarios, thereby driving the continuous landing of AI large models in many vertical industries such as finance, education, law, and medical care, and the demand ceiling for AI2B and AI2C is also expected to be fully opened.
Achieve breakthroughs in the high-quality application of large-scale model technology, and drive the "butterfly change" and industrial transformation of the real economy.
It is worth noting that in 2024, multiple favorable factors will promote the rapid development of large models. The first is the strong support from the government level, such as the "artificial intelligence+" initiative. The second is the surge in demand from users to improve their lives and work efficiency, coupled with the fact that technology companies have increased their investment in the field of AI, manpower, and technology research and development, and all links have coordinated to support the development of large models.
Among them, Beijing, Shenzhen, Hangzhou, Chengdu, Fujian, Anhui, Shanghai, Guangdong and other places have issued relevant policies on AI large models to accelerate the sustainable development of the large model industry. Taking Shanghai as an example, Pudong Zhangjiang, Xuhui West Bank, Lingang New Area, Minhang Maqiao and other places have formed four major industrial patterns, becoming an important carrier of artificial intelligence innovation and development, the number of AI enterprises above designated size has reached 350, almost doubled in the past five years, the output value has reached more than 380 billion yuan, almost three times that of five years ago, and the number of industrial talents has reached 230,000, accounting for about one-third of the country.
At present, general models, industry models, and end-side models have sprung up, and the application of the large model industry will be further accelerated. As the core driving force of the new generation of artificial intelligence industry, the AI model has widely empowered many fields of the mainland's economy and society, opened the door to general artificial intelligence, and promoted a new round of scientific and technological revolution and industrial transformation.
Domestic large models are promising
There is no doubt that the mainland AI large model industry is booming. With the launch of AI large models by many technology manufacturers, various general, industry, and end-side large models have achieved remarkable results in many fields, and AI large models have become an important means to improve service quality and efficiency in the fields of finance, medical care, and government affairs.
The representative general-purpose AI models in mainland China mainly include Baidu's Wenxin Yiyan model, iFLYTEK's iFLYTEK Xinghuo cognitive model, Alibaba's Tongyi Qianwen model, etc., and the industry AI model mainly covers Yonyou's YonGPT model and Ronglianyun's Chitu model. At the same time, the device-cloud combined AI large model with cloud-side and device-side large models mainly includes Vivo's blue heart large model, and the device-side AI large model is mainly represented by NIO's NOMI GPT large model.
In addition to the major Internet manufacturers have invested resources in vigorous development, operators are also gradually laying out the "large model" application market. China Unicom's "Honghu Graphic Model 1.0" is the first AI model for operators' value-added services. China Telecom has launched the large language model TeleChat, the "Xingchen" model with 100 billion parameters, and the first batch of 12 industry models have been put into trial use, basically covering all aspects of people's life, work and production. China Mobile released the "Nine Days" model, including the Nine Days, Haisu Government Affairs Model, and the Nine Days, Customer Service Model.
However, from the perspective of development, the application of the consumer side is only the tip of the iceberg of the AI model, and the real value of the large model is the application of the industry side and the enterprise side. In recent years, with the rapid development of artificial intelligence technology, industry models have gradually become an important driving force for the transformation and upgrading of various industries.
Compared with the general large model represented by ChatGPT, the industry large model can better understand the semantics and norms of the industry and perform more specialized tasks more effectively by training and optimizing the data in vertical segments.
As a general technology, the industry model can be widely used in finance, healthcare, education, retail and other fields to provide intelligent solutions for the industry. At the same time, it also gives the new thing of "large model" a foundation for landing, showing the potential of "AI empowering thousands of industries".
Regarding the future trend of AI large models, Zhou Hongyi, founder of 360 Group, said that 2024 is the first year of large model application scenarios, and China can completely embark on a road of large model development with Chinese characteristics.
Opportunities and challenges
As an important technology in the current field of artificial intelligence, AI large models have shown great potential in giving birth to new industries, new models and new kinetic energy, which not only supports the high-quality development of the economy and society, but also meets the requirements of innovation-driven and industrial upgrading emphasized in the Outline of the National Innovation-Driven Development Strategy, and is an important means to achieve the development of new quality productivity, which can promote intelligent upgrading in multiple fields, improve production efficiency, reduce production costs, and enhance industrial competitiveness.
In the future, driven by the strong drive of "supercomputing center + data resources + market demand", with the gradual innovation of the underlying technology, the technical route of "underlying technical support - training model and repeated training - scenario application - vertical industry model application" of the AI large model will be further clarified, and the general basic large model and vertical industry large model will continue to improve. It is worth mentioning that large AI models still face problems in computing power, algorithms, data security, and ethics.
First of all, the challenge of increasing the computing power demand of algorithms, along with the improvement of large model parameters and data, its requirements for computing power are also increasing. Second, the cost of data acquisition and management continues to increase. Finally, the ethics and credibility of large models need to be improved, and large models have bias and fairness problems, because the training data of the model may contain potential bias, which may lead to unfair results of the model, and large model development and application enterprises have also recognized the importance of governance.
Experts suggest that in the future, the mainland needs to further strengthen the overall planning of resources and R&D forces, strengthen the role of large models in the development of scenarios, and promote high-quality economic and social development, so as to achieve high-quality application breakthroughs in large model technology and drive the "butterfly change" and industrial transformation of the real economy.
Written by: Hu Yuan
Chart: Dawn
Editor and proofreader: Hu Yuan
Guidance: Xin Wen
Telecommunications Day Special Article| Promoting Sustainable Development: A World with Light, Without the Digital Divide Telecommunication Day Special Article Bridging the Digital Divide: 5G Takes Connectivity from "Wireless" to "Infinite"
Telecommunications Day Editor's Choice: Digital Innovation for Sustainable Development, Solutions and Typical Applications
What they say: Message from Industry Leaders on World Telecommunication Day (2)
Telecom Day Message | Hengtong Cui Wei: Develop new quality productivity and work together to move towards a better future
Telecommunications Day Special Article| 5G-A sets sail to boost the growth of new quality productivity
Telecommunications Day Special Article| Boosting Digital Innovation: China Information Technology 5G Intelligent O&M Empowers Smart Wireless Signal Day | Enabling new digital productivity, SIEMIC's next-generation all-optical access network helps sustainable development
Latest: The theme of this year's World Telecommunication Day has been determined, and the theme publicity of the World Telecommunication Day on May 17 and the industry research of "Telecom Day Editor's Choice" have been launched