Shulou(Shulou.com)11/24 Report--

On November 19, APUS Chairman and CEO Li Tao was invited to attend the sixth Tsinghua Conference and delivered a keynote speech on "grasping the initiative of AI model development and leading intelligence to good" in the branch lecture on "Scientific and technological Innovation and Enterprise Development". Through an insight into the development challenges of the global AI industry, Li Tao proposes to upgrade and update the "six elements of AI" to drive the sustainable development of China's AI industry, bring professional insight to the global AI construction and governance, and explain how to lead intelligence to the good.

This time, APUS stands on the world academic plateau to discuss the development and governance of global artificial intelligence with industry leaders, which is highly recognized by the industry for APUS's technology accumulation, scene practice and ecological construction in the field of AI. Li Tao's speech was also highly praised by the experts at the meeting.

Han Jingyang, vice chairman of Tsinghua University and former deputy secretary of the party committee of Tsinghua University, commented: Mr. Li Tao sharply pointed out that OpenAI's control over developers and users will bring great risks, which makes us feel the urgency of building our own country's AI general model base.

Zhou Xiaoyan, vice president of the China Association for the Promotion of International Investment, believes that the global competition around the dominance of AI starts from the struggle for the leading edge in science and technology. Mr. Li Tao talked about the far-reaching intention and pragmatic action of "intelligence for good" in his speech, which is precisely the purpose of intelligent development advocated by China's plan, which brings Chinese thinking to global AI governance.

High-level strategic dialogue brings together world wisdom

"Tsinghua Society" is an important academic brand created by Tsinghua University, which brings together elites in the field of politics and business at home and abroad, and used to discuss global governance and the construction of a community with a shared future for mankind along the guidance of science and technology. The event has been held for five consecutive years and has become a high-level platform for strategic dialogue and an active think tank in academic circles.

This year, with the theme of "Mutual Learning among Civilizations and peaceful Development", Tsinghua University will invite global "think tanks" to talk freely about new knowledge around the three thematic chapters of "Mutual Learning among Civilizations and peaceful Development", "Scientific and technological change and Global Governance" and "Scientific and technological Innovation and Enterprise Development." it has attracted wide attention from the international community.

Tsinghua University attached great importance to the event. Qiu Yong, secretary of the party committee of Tsinghua University and chairman of Tsinghua University, delivered a speech at the opening ceremony. Peng Gang, vice president of Tsinghua University and vice chairman of Tsinghua University, presided over the opening ceremony. Hu Xianzhang, former deputy secretary of the party committee of Tsinghua University and vice chairman of the Tsinghua speech Advisory Committee, presided over the closing ceremony, and Han Jingyang made a concluding speech.

Ambassadors to China from South Korea, Indonesia, South Sudan, Finland and Slovenia, dozens of well-known experts and scholars from the Party School of the CPC Central Committee, the Institute of Science and Technology Strategic Consulting and Research Institute of the Chinese Academy of Sciences, Seoul University, Nankai University and think tanks gathered in Beijing to conduct multi-disciplinary speeches and discussions on themes, so as to contribute academic wisdom and strength to the implementation of the Global Development Initiative and the Global Civilization Initiative.

Global AI governance receives high attention

At present, a new round of scientific and technological revolution and industrial reform are in the development period of achieving major breakthroughs. The rapid development of cutting-edge science and technology has aroused global concern and controversy about the risks of science and technology ethics, national security, citizen privacy and data security. As the key factor of science and technology game and strategic decision of great powers, how to coordinate the development, security and governance of artificial intelligence has become the focus of the world.

The meeting discussed in depth the construction of scientific and technological governance system and the building of scientific and technological governance capacity, and advocated that through international cooperation, we should control the force of scientific and technological change and promote the establishment of a more just, reasonable and balanced global governance system. In the course of the lecture, a number of Tai Dou-level experts and academicians and industry leaders focused on what artificial intelligence should govern and how to govern, giving insights into the direction of industrial development.

Gong Ke, former president of the World Federation of Engineering organizations and executive director of the China New Generation AI Development Strategy Research Institute, pointed out the five major governance directions of artificial intelligence from the macro level, namely, "AI development-oriented governance, based on ethical governance, risk classification and hierarchical governance, full participation in governance, and global pluralistic co-governance pattern". He mentioned that it is necessary to accelerate the development of standards conducive to the development of AI productivity and form a consensus on global governance.

Zhang cymbal, professor of the Department of computer Science of Tsinghua University, academician of the Chinese Academy of Sciences, and winner of the 2019 highest Achievement Award for artificial Intelligence by Wu Wenjun of the Chinese Society of artificial Intelligence, pointed out the good governance strategy from the technical level.

Zhang cymbal believes that security issues can focus on "the model itself and the user's NPU", these two different nature of security issues, governance methods should also be different. He pointed out that the AI large model is the most important achievement in the development of artificial intelligence at present, and we should pay special attention to the security of the results generated by the AI large model, and pay special attention to the fragility of the model itself caused by the training data. AI alignment and building a standard constraint large model is an efficient and fast governance strategy.

Over the past few days, China has issued the "Global artificial Intelligence Governance Initiative", put forward the Chinese plan for artificial intelligence governance, and established the purpose of "intelligence for good"; the first consensus reached at the high-level meeting between the two major global economies is the intergovernmental dialogue on artificial intelligence; the upgrading of OpenAI has caused a shock in the industry, and the global development and governance of artificial intelligence have ushered in new changes.

From the point of view of the provider of the large model base, Li Tao deeply analyzed the global potential ecological risks of OpenAI from the perspective of AI upgrade, and proposed to guard against the "AI empire" to carry out "three monopolies" of data, traffic, and technology, resulting in heavy user dependence, developers losing their brands, resulting in unhealthy industrial ecology, aggravating the intelligence gap, and challenging the six major impacts of the current government, society, and global operating rules in the future.

Li Tao believes that the international community needs to reach a consensus: all major economies should build their own AI universal large model base; have a set of AI framework or AI AGENT in line with their own national values to contain and restrict the large model; and actively build a healthy artificial intelligence industry ecology to resolve the global AI risk challenges.

Focus on six elements and grasp the initiative of AI development in China

A few days ago, the Ministry of National Security released "how to resolve the National Security Challenge brought by artificial Intelligence", pointing out that it is necessary to continuously promote the modernization of the national security system and capability. to build a solid national security barrier to deal with the risk challenges of artificial intelligence-it is increasingly urgent to promote the development of artificial intelligence in a responsible manner.

Li Tao believes that as the second largest economy in the world, China must have its own AI general model base; to build a red corpus, to align the values of Chinese models, to contain and restrain them through the CHINA AI AGENT (CAA) mechanism; and to build a healthy ecology of China's artificial intelligence industry to ensure the health and stability of China's artificial intelligence, and to move forward in line with China's mainstream values and the evolution of civilization.

Li Tao succinctly adapts to the key "six elements" of China's AI development, helping China to grasp the initiative in the development of AI models and leading intelligence to the better. They are robust computing power, global knowledge and high-quality data, agile evolutionary algorithms, values alignment, healthy industrial ecology and value creation.

Li Tao said: only by building surging computing power can we meet the needs of complex scenes, achieve agile training and high-speed parallelism, and support the resonance of large models with the same frequency of the times; through the collection of global knowledge and high-quality data training, can make Chinese large models have higher performance; with the real market environment as the only test standard, efficient iterative model algorithm is to explore a "better solution" to reduce cost and increase efficiency of large models.

In addition, the Chinese model must pay attention to values alignment and layout, in order to face different scenarios, and provide users and industries with positive and high-quality content and applications. As for the potential global ecological risks, Li Tao believes that reasonable and scientific division of labor and distribution of interests can create a healthy industrial ecology and drive the sustainable development of the AI industry.

The responsibility of enterprises under the action of a great power

Tsinghua Society believes that in the great era of change, Chinese scientific and technological enterprises need to further speed up the improvement of industrial innovation chain, build collaborative innovation platform, create innovation ecology, and enhance the systematic capacity of scientific and technological innovation. in order to build a deep integration of industry, university and research innovation system, and open up the channel from strong science and technology to strong industry, strong economy and strong country.

At present, the application of AI large model has moved towards the depth of industry. Chinese scientific and technological innovation enterprises, especially artificial intelligence enterprises, bear the important task of leading the model value and promoting pragmatic innovation.

APUS is committed to building a big AI model for China, bringing the application of the big model into line with value creation. Since the release of APUS model, it has realized multi-industry energy by virtue of the encapsulation and application of deep understanding and generation ability of text, image, audio and video. Based on industry vertical data distillation, it has extracted APUS model, APUS e-commerce model, APUS network model, APUS education model, APUS manufacturing model and other industry models, and efficiently implemented AI applications.

APUS attaches great importance to values alignment and adheres to industry self-discipline. The APUS model has accumulated a global high-quality corpus based on 2.4 billion users, while working with the industry, together with the national big data platform and BAT, to carry out technical research, industrial cooperation and industry self-discipline around artificial intelligence security and governance.

In addition, APUS pays attention to close cooperation with all aspects of industrial ecology, with consultation, co-building, co-governance, sharing and win-win as the traction, and on constructing a complete panoramic structure map of AI industry in accordance with the "six elements of AI", striving to build a healthy industrial ecology, opening up API interfaces and open source models to Chinese developers, helping to achieve universal benefits of AI and bridging the intelligence gap.

Artificial intelligence will be the core technology that will affect the next century. In order to develop artificial intelligence well, we need to take a good step, lay out a good situation, firmly grasp the initiative in the development of AI model, and let Chinese applications grow on China's AI model. This is not only the core fortress of China's AI development, but also the development essentials of leading the "variable" to a good way. APUS will play an active role in building a credible and reliable innovative application base for the development of AI in China.

The lecture will be sponsored by Tsinghua University, co-sponsored by the National Governance and Global Governance Institute of Tsinghua University, Belt and Road Initiative Strategic Research Institute of Tsinghua University, School of Social Sciences of Tsinghua University, and co-sponsored by the Silk Road International think Tank Exchange Center and the Silk Road Youth Forum. The event is supported by media such as Guangming Network, Xinhuanet and so on.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.