BEIJING, Beijing, Oct. 26 (Xinhua)-It’s a familiar poem that apes say goodbye to each other. Man is the spirit of all things, benefiting from and growing up in nature, creating and using different tools to face vicissitudes of life, and only then can there be a great beauty room on which to live today. Today, everyone will regard artificial intelligence technology as the strongest innovative application cultivated by human beings. Indeed, artificial intelligence is leading a new round of industrial transformation and bringing great development opportunities to the world, but at the same time it has also triggered a series of new risks and challenges, and may even pose a devastating threat to the future survival of mankind. Whether we live in the "human world" or in the "artificial intelligence room" is a human problem. All countries are talking about artificial intelligence ethics, but what will happen in the future if it goes against the basic rule of "people-oriented", the development and application tool of human society? Exploring reasonable and effective governance schemes to realize the healthy and orderly development of artificial intelligence has become an important issue in the world. At the third "Belt and Road" international cooperation summit forum, China released the Global Artificial Intelligence Governance Initiative (hereinafter referred to as the Initiative), which won wide acclaim from the international community. Standing at the historical height of building a community of human destiny and safeguarding the well-being of all mankind, the Initiative puts forward constructive solutions in three aspects of artificial intelligence development, security and governance fairly and fairly, and provides a Chinese solution for global artificial intelligence governance.
First, build a safe bottom line and effectively respond to the general concerns of the international community.
As China’s centralized response to the global artificial intelligence governance, the Initiative emphasizes security controllability, privacy protection, fairness and non-discrimination, and gives countermeasures and methods based on the actual situation of related issues and the governance work in China, which effectively responds to the issues that the international community is most concerned about on the safety of artificial intelligence and provides a blueprint for all parties to carry out the safety governance of artificial intelligence. Among them, the core problem of safety and controllability is to improve the interpretability and predictability of artificial intelligence and ensure that human beings can trust the behavior of machines, which is a prerequisite for human beings to promote the development of artificial intelligence. In practice, with the wide application of deep learning and other technologies, the self-iterative ability of artificial intelligence has been continuously enhanced, and its dependence on human beings has been continuously reduced in the process of training and training; The more complex internal network structure also makes the black box attribute of artificial intelligence technology continuously enhanced, and the balance of human trust in the output results of artificial intelligence by mastering its own operating rules becomes more fragile. The relatively opaque artificial intelligence will make it difficult to guarantee the security and fairness of its related data processing behavior, and the accountability mechanism will also be affected in some complex applications. Based on reality, the Initiative takes improving interpretability and predictability as the key work of artificial intelligence research and development, ensures the authenticity and accuracy of relevant data, and puts forward practical and feasible measures for the safety management of artificial intelligence. In practice, focusing on improving the transparency and reliability of artificial intelligence data sources, system management and other fields based on this proposition will not only help to realize auditable, supervised, traceable and reliable artificial intelligence technology,It will also better meet the needs of data and personal information security, fairness and non-discrimination in the field of artificial intelligence on the premise of enhancing human trust in artificial intelligence.
Second, focus on the governance mechanism and actively promote the cooperation framework of joint construction.
After recognizing many risks and difficulties in the development and application of artificial intelligence, the international community has begun to actively explore and promote the work related to artificial intelligence governance. At present, the initiatives, guidelines and suggestions issued by governments, international organizations, enterprises and people from all walks of life have their own emphases, and there is still a significant distance from the formation of a unified rule system and governance mechanism recognized by the international community. Based on this, the Initiative issued the idea of collaborative governance and the establishment of an international artificial intelligence governance institution. It is hoped that by coordinating countries to strengthen communication and dialogue, multi-stakeholders will be promoted to participate in governance matters in depth, and major issues related to the development, safety and governance of artificial intelligence will be coordinated, and the global artificial intelligence governance process will be continuously promoted. The above claims can directly hit the two major pain points of the current artificial intelligence governance work. First, different countries have different understanding, development, application ideas and methods of artificial intelligence technology. For a long time, they have formed an artificial intelligence ecosystem with different levels and characteristics, with different emphasis capabilities and development directions. Therefore, there are also many differences in the artificial intelligence governance strategies designed. By coordinating the communication mechanism and sharing the best operational practices, it can not only promote all parties to achieve a broad consensus on artificial intelligence governance, but also form universal theoretical rules that reflect the wishes and respect the interests of all parties, and also brainstorm practical measures to effectively prevent and deal with artificial intelligence risks, greatly improving the overall international level of artificial intelligence governance. Second, the huge risk of uncontrollable artificial intelligence technology has become a key issue for the international community to carry out artificial intelligence governance.Therefore, it is inevitable to adhere to multilateralism, set up international governance institutions and discuss risk coping strategies. Considering the great influence of artificial intelligence on today’s society, we can take the establishment of nuclear energy governance mechanism as a reference. With the establishment of the International Atomic Energy Agency and the promulgation of the Nuclear Non-Proliferation Treaty and other documents, through decades of long-term coordination and cooperation among all parties in the international community, the application of nuclear weapons has gradually shifted from military confrontation to social welfare, contributing to the well-being of mankind. We can learn from relevant work experience, establish an international artificial intelligence governance organization based on the "Initiative", advocate the establishment of an artificial intelligence governance cooperation framework, promote mutual trust and mutual trust, and effectively achieve the balance between artificial intelligence development and security in the future.
Third, clarify the principles of development, promote artificial intelligence to enhance human welfare
Due to the differences in geography, culture, economy, productivity and other factors, countries and regions in the world have different production forms, lifestyles and group values, and there are also obvious differences in the cognition and usage habits of a certain technology and tool among different civilizations and governments, resulting in new difficulties and challenges. The development of artificial intelligence is also facing this situation. With the development of technology, risks and challenges are gradually increasing. On the one hand, artificial intelligence technology is constantly evolving due to its integration with algorithms and data, but the risk of technology-related uncertainty also increases in the process of evolution. Some technical achievements developed for specific scenarios have not yet formed effective governance ideas, even the most developed and active countries and economies today cannot guarantee that artificial intelligence is fully understood, mastered and controlled by human beings, and effectively control the risks and threats caused by technology; On the other hand, a digital divide and a technological divide have emerged in the field of artificial intelligence. Some developed countries make use of their first-Mover advantages in technological innovation, product research and development, ecological creation and other fields to use artificial intelligence at will, and some groups even directly use it to manipulate public opinion, interfere in other countries’ internal affairs, infringe on their sovereignty, threaten the security of cyberspace, and endanger people’s peace of life and the stability of international order.
Faced with the above problems, it is necessary to form a unified concept and understanding from the perspective of all mankind, design the governance rules and order to be followed in the development and application of technology in the era of digital, network and intelligence with common values, and form a global consensus so as to coordinate and discuss the controllable development direction of artificial intelligence more effectively. To this end, from the perspective of respecting human rights and ensuring human security, the "Initiative" puts forward the proposition of people-oriented, intelligent goodness and respect for sovereignty, clarifies the relationship between artificial intelligence and human beings, and gives basic principles for the future development direction of artificial intelligence. Among them, the central idea of people-oriented and intelligence-oriented is that the development and utilization of artificial intelligence technology should adhere to the basic principle of promoting the growth and progress of human society and enhancing the common well-being of mankind; Respect for sovereignty focuses on the development and application practice of artificial intelligence technology, emphasizing that the development of artificial intelligence should respect the global order and safeguard the sovereignty of all countries. The "Initiative" will play a great role in promoting the development and utilization of artificial intelligence by countries and institutions around the world, preventing malicious organizations and lawless elements from abusing artificial intelligence technology, and ensuring that artificial intelligence technology cannot become a weapon and tool for attacking human beings. (Author: Li Yuxiao Secretary General of China Cyberspace Security Association)
关于作者