HyperAI超神经

This morning (March 5), the third session of the 14th National People's Congress officially opened, which coincides with the final year of my country's "14th Five-Year Plan". Standing at this important node with connecting the past and the future, all sectors of society are paying more and more attention to this year's "Two Sessions" - how can the policy side further exert efforts in 2025 to achieve the goals of the "14th Five-Year Plan" and lay a solid foundation for the start of the "15th Five-Year Plan"?

Looking back at 2024, the Premier pointed out in the Government Work Report that "the main goals and tasks of economic and social development for the whole year were successfully completed, high-quality development was steadily promoted, new quality productivity developed steadily, and my country's economic strength, scientific and technological strength, and comprehensive national strength continued to increase." At the same time, my country's new achievements in the fields of integrated circuits, artificial intelligence, quantum computing, etc. have frequently emerged, which further demonstrated the improvement of innovation capabilities under the construction of a strong country in science and technology.

It is worth mentioning that "AI+" was first proposed in the 2024 Government Work Report.We can see that from the digital and intelligent transformation of all aspects of food, clothing, housing and transportation to breakthroughs in large models, robots, autonomous driving and other fields, in this first year of "artificial intelligence +", domestic AI development has achieved gratifying results.

During the just-passed Spring Festival, DeepSeek set off a wave of breakthroughs in domestic AI and also became a hot topic at the two sessions.

At the National People's Congress press conference on March 4, Lou Qinjian, spokesman for the conference, said in response to a reporter's question that the significant progress made by DeepSeek represents the rise of a group of Chinese companies in the field of artificial intelligence. The creative team composed of young and middle-aged people is taking the lead in the advancement of China's science and technology, which deserves praise. He also said,"Through the rise of companies like DeepSeek, we can see China's innovation and inclusiveness in technological development."

At the first "Minister's Channel" interview on March 5, Minister of Education Huai Jinpeng said that DeepSeek and robots have attracted widespread attention at home and abroad in recent times, which also illustrates the effectiveness of China's scientific and technological innovation and talent training.

Looking ahead to 2025, technology will continue to be one of the key topics.In the Government Work Report, the Premier proposed to continue to promote the "AI+" action.Better combine digital technology with manufacturing advantages and market advantages, support the widespread application of large models, and vigorously develop a new generation of intelligent terminals and intelligent manufacturing equipment such as intelligent networked new energy vehicles, artificial intelligence mobile phones and computers, and intelligent robots. Expand the large-scale application of 5G, accelerate the innovative development of the industrial Internet, optimize the layout of computing resources across the country, and create a digital industry cluster with international competitiveness. Accelerate the improvement of data infrastructure systems, deepen the development and utilization of data resources, and promote and standardize cross-border data flows. Promote the standardized and healthy development of the platform economy, and better play its positive role in promoting innovation, expanding consumption, and stabilizing employment.

In terms of scientific research, the government work report proposes:Promote science and technology spending to tilt towards basic research,Improve the investment mechanism that combines competitive support with stable support, and improve the degree of organization of basic research. Give full play to the leading role of science and technology leading enterprises, strengthen the deep integration of industry, academia and research led by enterprises, and institutionally guarantee that enterprises participate in national scientific and technological innovation decisions and undertake major scientific and technological projects.

In terms of the development of emerging industries, the government work report proposes to establish a mechanism for the growth of future industry investment, cultivate future industries such as biomanufacturing, quantum technology, embodied intelligence, and 6G, deepen the pilot program of integrated development of advanced manufacturing and modern service industries, and accelerate the development of service-oriented manufacturing.

Today is also the "Qingzhe" in the 24 solar terms. When spring thunders and everything sprouts, it also implies that my country's science and technology industry is accumulating strength and will eventually make great progress. As "warriors" on the front line of science and technology development, the helmsmen of domestic technology companies also have their own insights on the development trends of the industry. During this year's two sessions,Lei Jun, Zhou Hongyi, Liu Qingfeng and other bigwigs continued to actively offer their opinions and suggestions, and the topic was closely related to AI.

Liu Qingfeng: Systematically build "AI + Medical" standards and specifications to promote the healthy development of the industry

Liu Qingfeng, deputy to the National People's Congress and chairman of iFLYTEK, put forward nine suggestions on autonomous and controllable ecological construction, talent cultivation, employment security, AI + medical care, AI + education, smart terminals, technology for the elderly, and technology for the disabled in the AI era, which are:

* Accelerate the construction of independent and controllable large models and industrial ecology on domestic computing platforms

* Prevent the harm caused by large models generating “illusion data” that floods the Internet, and build a credible information environment from the perspective of technology research and development and management mechanisms

* Optimize the per capita funding structure to provide continuous funding support for "AI + Education"

* Systematically build a talent profile and training system in the AI era, and innovate evaluation methods

* Develop new AI jobs, improve AI unemployment insurance, and create an AI employment-friendly society

* Enrich and optimize policies such as "trade-in" to promote the development of AI smart terminal industry

* Systematically build "AI + Medical" standards and specifications to promote the healthy development of the industry

* Accelerate the construction of a technology-friendly system for the elderly, and help the elderly share the "AI dividend"

*Use technology to help the disabled create a barrier-free social environment and innovate to build a humanistic care ecosystem

Among them, in the construction of "AI + Medical",Liu Qingfeng said that big model technology is widely used in medical clinical, service, management and other fields, with high professional requirements and fast technology implementation. my country currently still lacks systematic standards and specifications for "AI + medical care", and faces problems such as data security, medical professionalism and ethical risks. For example, some AI applications do not have professionally certified medical capabilities but can answer related questions, causing great cognitive misunderstandings.

In this regard, he suggested formulating technical standards and testing specifications for the application of big models in the medical and health industry: focusing on testing the clinical practice value of big model technology innovation, identifying authoritative third-party evaluation agencies and information disclosure mechanisms, selecting qualified regions and technology companies to conduct trials first, and formulating application technology standards and testing specifications.

Lei Jun: Strengthen the governance of the hardest-hit areas of "AI face-changing and voice-changing" illegal infringement

Lei Jun, deputy to the National People's Congress and chairman and CEO of Xiaomi Group, brought five proposals, namely:

* Accelerate the mass production of autonomous driving

* Develop the intelligent connected new energy vehicle industry ecosystem

* Accelerate the high-quality development of the artificial intelligence terminal industry

* Optimize the design of new energy vehicle license plates

* Strengthen the governance of areas hit hard by illegal infringement of "AI face-changing and voice-changing"

Among them, regarding the illegal infringement management of "AI face-changing and voice-changing",Lei Jun suggested speeding up the process of legislation of individual laws, raising the level of legislation on the basis of giving equal importance to safety and development, strengthening industry self-discipline and co-governance, and consolidating the responsibilities and obligations of platform companies and other parties. Increasing the breadth and intensity of legal publicity, and enhancing the public's vigilance and discernment.

Zhou Hongyi: Still focusing on AI and security

Zhou Hongyi, member of the National Committee of the Chinese People's Political Consultative Conference and founder of 360 Group, attended the two sessions for the eighth year and brought three proposals, namely:

* Suggestions on comprehensively solving the security issues in large model applications

* Suggestions on moderate tolerance of large model "illusions"

* Solve the dilemma of the cybersecurity industry with security as a service

Regarding the large model illusion problem,He believes that "hallucination" is a double-edged sword. It is an innate characteristic of large models that cannot be eliminated. It also determines their creativity and imagination. It is extremely important for achieving AGI (general artificial intelligence) and can shine in scientific research, literary and artistic creation and other fields. In fields such as medicine, law, and finance that require high accuracy, although "hallucination" may be wrong, it can be corrected by using technical means, such as RAG (retrieval enhanced generation) technology, by comparing professional knowledge bases, real-time network information, etc.

In this regard, he suggested that the development of large models is related to the country's destiny, and its creativity should not be restricted due to excessive pursuit of accuracy. It is necessary to give flexible supervision to "illusions" and create a good environment that supports innovation and tolerates mistakes, so as to allow more companies to "copy" DeepSeek's success.It is suggested to learn from the successful experience of the "safe harbor principle" and implement flexible supervision on big models and related products and services, and moderately tolerate the "illusion" of big models.For general mistakes caused by "illusions", avoid direct shutdowns and removals, give companies the opportunity to correct themselves, and encourage them to innovate boldly and compete freely.

Qi Xiangdong: It is recommended to issue documents such as mandatory compliance requirements for large-scale model network data security

The proposal of Qi Xiangdong, member of the National Committee of the Chinese People's Political Consultative Conference and chairman of Qi'anxin Technology Group Co., Ltd., focuses on the security governance of artificial intelligence.In terms of policy, he suggested issuing documents such as mandatory compliance requirements for large-model network data security to provide clear guidance for companies to do a good job in security protection in the era of artificial intelligence; encouraging the industry to conduct regular network and data security "physical examinations" to help companies check for gaps in security capabilities and achieve continuous improvement in security capabilities; setting up special funds to promote the implementation of "AI + Security" innovation results, promote joint innovation between leading companies in various fields and professional network security companies, and improve the security protection effectiveness of emerging scenarios such as smart cities, smart energy, and smart finance.

Zhou Yunjie: It is recommended to use industrial big models as key variables to enable new industrialization

National People's Congress Deputies,Zhou Yunjie, Chairman and CEO of Haier Group, proposed the suggestion of "using the industrial big model as the key variable to enable new industrialization".He said that the first step is to consolidate the development foundation and publish national industrial scene maps, corpora and data sets. He suggested setting up a national special project and intensifying efforts to promote and implement the "one map and four lists" action plan to provide a solid foundation for the application of artificial intelligence in the industrial field.

Second, support leading enterprises and demonstrate the application value of industrial large models. It is recommended to support leading enterprises in key industries such as home appliances, automobiles, and medical care through financial subsidies to carry out pilot trials of deep application of artificial intelligence based on industrial large models, create industry benchmarks, and demonstrate application value; through financial subsidies, encourage leading enterprises to refine and share typical cases of deep application of artificial intelligence in key links such as R&D design, production and manufacturing, and business management.

The third is to support platform enterprises and promote industrial large models to serve small and medium-sized enterprises. It is recommended to introduce special financial support policies for national-level cross-platform enterprises, cultivate a batch of standardized, reusable, and low-cost solutions that accurately match industrial large models with scene graphs, and provide small and medium-sized enterprises with useful, affordable, and effective artificial intelligence services.

Cao Peng: Strengthening heterogeneous computing power construction to achieve intelligent computing for all

Cao Peng, member of the National Committee of the Chinese People's Political Consultative Conference, Chairman of the JD Group Technology Committee and President of JD Cloud Division, also submitted a number of proposals during this year's two sessions, including vigorously cultivating new physical enterprises to promote the deep integration of scientific and technological innovation and industrial innovation; strengthening the construction of heterogeneous computing power to achieve intelligent computing and inclusive services for the development of new quality productivity; supporting private enterprises to strengthen the protection of the rights and interests of new employment groups;Use AI technology to promote the popularization of Internet medical care and cultivate new quality productivity of digital medical care;Suggestions include strengthening the integrated construction of supply chain finance, etc.