[Expert interpretation]

Author: Gao Wen (member of the National Science and Technology Ethics Committee, academician of the Chinese Academy of Engineering)

The "Opinions on Strengthening Science and Technology Ethical Governance" (referred to as the "Opinions") is a landmark programmatic document in the field of science and technology ethics governance in my country in the new era. It provides basic guidelines for promoting the development of "science and technology for good", serving my country's social progress and benefiting mankind. It is of far-reaching significance for effectively preventing science and technology ethical risks and promoting the modernization of science and technology governance systems and governance capabilities.

The research content of artificial intelligence often involves sensitive areas of scientific and technological ethics. According to the spirit of the "Opinions", I think the following work should be mainly carried out:

One is the moral level. In accordance with the spirit of the "Opinions", it is necessary to improve scientific and technological ethics norms and guidelines in the field of artificial intelligence, to provide ethical and moral restraint and guidance to scientific and technological personnel, so that scientific researchers have a better sense of ethics and morals, and always constrain themselves with ethical and moral red lines during the research and development process. Promote the establishment of ethics committees in artificial intelligence research institutes and enterprises, consciously carry out artificial intelligence ethical risk assessment, monitoring and real-time response, so that moral guidance and moral constraints run through the entire process of artificial intelligence design, research and development and application.

The second is the technical level. It is necessary to strengthen the technical control and prediction capabilities of front-line scientific and technological personnel so that they can promptly identify the development stage and extent of various potential risks (for example: risks are not currently serious, but potential risks are definitely there; the immediate urgency of risks is not strong, but the long-term severity must not be underestimated; the severity of risks is generally underestimated, and there is no effective supervision and control, etc.). At the same time, it is necessary to strengthen the management, detection and evaluation of algorithms, data and applications, build an effective risk warning mechanism, and reduce ethical risks through continuous improvement of technology.

The third is the legal aspect. It is necessary to put the construction of laws and regulations in an important position and use them as the highest basis for ethical governance of artificial intelligence. We must continuously improve the legal system for ethical governance of artificial intelligence so that the entire process of artificial intelligence technology from design, research and development, testing, industrialization to application is legally based, and eliminate the blind spots of legal supervision of ethical governance of artificial intelligence.

Artificial intelligence research work should be guided by the "Opinions" and accurately grasp the requirements of artificial intelligence governance, which should not only promote the rapid development of artificial intelligence technology, but also ensure that artificial intelligence technology complies with social ethics.

First of all, we must firmly establish the awareness of moral and ethical red lines and draw the bottom line of ethical governance. Technological development often precedes ethics, so ethics must always be regarded as a prerequisite for scientific and technological innovation and technology application. Especially before entering unknown areas for exploration or pioneering applications, potential moral and ethical risks must be comprehensively evaluated, and the degree of risk must be assessed through various forms such as thought experiments and ethical paradox deductions, and potential risk points that may trigger moral ethics must be identified. Scientists must always regard basic human morality and ethics as an insurmountable red line, form strict moral and ethical boundaries and bottom lines in scientific research activities, and engage in various scientific research innovation and application practice activities within the moral and ethical safety zone.

Secondly, it is necessary to establish an agile response mechanism to actively respond to problems that arise in science and technology ethics governance. In the practice of scientific and technological activities, humans will face many new and unforeseen ethical governance challenges. On the one hand, we must insist on putting ethics first; on the other hand, we must establish a complete set of agile response mechanisms to quickly deal with emerging moral and ethical risks, promptly organize experts in ethics, humanities and social sciences, law, science and technology policy and other science and technology researchers to jointly carry out assessment and research, and promptly issue industry conventions, scientific and technological ethics standards, laws and regulations, etc., to minimize the negative impact of ethical governance lagging behind scientific and technological development.

Finally, it is necessary to strengthen international exchanges and cooperation and explore a Chinese plan for ethical governance of science and technology. In terms of morals and ethics, different countries, different ethnic groups, and different cultures have similarities and similarities, but they also have their own unique perspectives and positionings. We must stand at the strategic height of a community with a shared future for mankind, adhere to the basic principles of scientific and technological ethics governance of "enhancing human welfare, respecting the rights of life, adhering to fairness and justice, rationally controlling risks, and maintaining openness and transparency", and strive to promote scientific and technological ethics governance plans that are in line with all mankind. At the same time, we must also combine our country’s five thousand years of cultural tradition to conduct in-depth research on our country’s unique scientific and technological ethics governance path. In particular, we must proceed from the interests of the overwhelming majority of the people and actively prevent a series of moral and ethical issues such as data security and privacy leaks, algorithm discrimination, security risks based on deep forgery, rule of law risks, and technological poverty. We must not only focus on solving the ethical governance issues faced by China itself, but also make China's contribution to the development of all mankind.

"Guangming Daily" (Page 16, March 24, 2022)