In the past 60 years, scientific research on artificial intelligence (AI) has evolved from simulating human intelligence to an enabling technology that advance human and social development. AI, a technology crucial to global digital advancement, is now the developmental priority of many countries and intergovernmental organizations. In the meantime, the impact of AI on science, technology and policy is profoundly influencing international relations and the global landscape.
However, some countries and intergovernmental organizations are concerned that their leading position in the field of AI will be challenged in an era of global development, or mistakenly assume that the development of AI is a zero-sum game. As a result, recent trends have shown that the scientific and technological exchanges between scholars and the industries in different geolocations are being negatively impacted. Some countries are even being tried to be isolated from others in the field of AI development.
AI promises wide applications in countries across the globe and in all walks of life. It is an important channel that establishes ties between different countries, enabling cultural communication and interactions. For this reason, countries across the globe are working hard on the formulation and implementation of strategies that would promote the development and application of AI related technologies. In recent years, China has steadily increased fundamental R&D capacity, continuously expanding potential applications of AI. This has opened up more opportunities and secured advantages in the international community. In this context, the State Council issued the “New Generation Artificial Intelligence Development Plan,” which laid out the groundwork for China to become a premier AI innovation center of the world by 2030. Yet it should be noted that no country could possibly become the sole center of global AI development. The global AI network of the future will feature a group of major centers for basic research, industrial R&D and application services. These premier research centers will feature high interconnection and intensive collaboration, which will jointly drive forward the global development of AI.
The application of AI will serve all humankind, influencing future social development and human destiny. Like the large quantity of research-based groundbreaking technologies that have influenced the future of humankind, the development of AI is not a zero-sum game. Take brain science as an example. It might still take hundreds of years for scientists to decode the brain’s structures and cognitive functions to reveal the nature of human intelligence. This is why neuroscience researchers across the world have joined hands to develop institutions like the International Brain Initiative (IBI) to coordinate the development of brain research projects in different countries. This organization has enabled cooperation and resource sharing. The Brain Observatories have been instituted, promoting global data sharing of brain research and scientific discoveries. It takes the world to decode the brain. Likewise, and it is even more essential to establish a mechanism for long-term international cooperation on AI to advance the development of humankind.
The principles of AI governance serve as an important bridge and fundamental policy for countries to carry out mutually aided AI development. In the “Governance Principles for the New Generation Artificial Intelligence — Developing Responsible Artificial Intelligence,” China proposed to promote global sustainable development through AI. China will do so by putting forward the principles of harmony and human-friendly, inclusion and sharing, openness and collaboration. These concepts represent China’s vision for the development of next-generation AI, and China’s policy for global AI cooperation.
Inclusive development is based on the recognition of diverse cultures and values, thus necessitating cultural interaction, mutual understanding, increased consensus building and managing differences against the backdrop of diverse cultures. For example, the concept and expression of “harmony” is widely used in Chinese, Japanese, South Korean and African philosophy and cultures. The understanding and development of “harmony” in different cultures enriches and deepens the implications of relevant concepts in modern society. This shared understanding is likely to inspire the harmonious coexistence between human beings and AI in the future. It is a shared goal of humankind to achieve sustainable development, irrespective of differences in values and cultures or regional and political divergences. All parts of the world should seek to realize environmentally sustainable, AI-enabled development of society and the economy. To this end, countries need to build on the basic research of AI to promote exchanges. The human society will benefit from sharing AI evaluations, application scenarios, best practices and the various efforts on technological, social and ethical risks, among others.
The core of fundamental AI research lies on innovations of models and algorithms. These are generally considered to be the strategic high ground of competition among countries. Nevertheless, even in this pivotal field, the adequate sharing of outcomes remains to be the best solution to speed up innovation. Regarding technical AI, challenges based on the same data and tasks, plus regular international symposiums and competitions, are important tools to foster virtuous competition. This allows AI researchers and engineers to inspire each other to come up with innovative ideas for the future, bringing forward better models and algorithms. Open-source algorithms and systems are the most effective tool to maximize the impact of innovation. Only in this way can we employ research and technology communities to push forward the rapid iteration of AI algorithms and systems. As a representative example, the extensive development and application of deep learning models and systems worldwide is driven by relevant open-source communities. These directly benefit creators of said algorithms and systems, bringing about the honor and recognition they deserve. As the history of technological development shows, community contributors from countries all over the world will be able to identify and repair loopholes and safety hazards lurking in algorithms immediately through open-source code and models. Open ecosystems are gradually taking shape in the field of AI, fueling the fast development of AI technologies. Through collaboration, AI innovators, policymakers and governments across the globe are obliged to carry forward and maintain these mechanisms and platforms for virtuous competition and development.
The current development of AI is still highly uncertain. Countries need to jointly cope with the technical safety, security and ethical issues arising out of its development. Otherwise, many regions will fall victim to the negative impacts of technological advancements, regardless of previous lessons learned in other countries and regions. For example, some middle schools and universities in China adopted an application of AI-enabled automatic facial expression recognition in classrooms. However, this infringed on the privacy of students, negatively impacting the interactions between students and teachers. Eventually, various education departments and members of the public spoke out against the system, resulting in its suspension. Unfortunately, these relevant experiences and responses were underappreciated by other countries. The United States and some European countries let similar cases repeat themselves in the following months. That being said, AI should not be completely banned in schools. The critical issue is whether we can protect the privacy of individuals while simultaneously employing technology to benefit all. For example, Japan has set a good precedent for other countries, as some Japanese primary and middle schools use AI to prevent school bullying. In addition, there are both long-standing technological and cultural risks lurking in different paths to realize AI. For example, AI is primarily considered to be a tool in the West, but Japan regards the technology as a companion and quasi-member of the society. In most science fictions, AI is often depicted as the common enemy of humankind. In the long run, different visions for the relationships between AI and humankind will generate different types of risks in the course of technological development. As such, it is of vital importance to enter into global collaboration to circumvent such risks, jointly examine long-term issues and forestall risks. On top of that, we should make efforts to prevent vicious competition, misuse, abuse and potential evil use and applications of AI against great uncertainties during the technological development.
China is dedicated to becoming one of the world’s premier AI innovation centers. On this account, we need to first clarify two specific considerations. To begin with, every center has different developmental concepts, advantages and priorities. Therefore, it is critical that they learn from each other. The US and United Kingdom boast prominent advantages with regards to being the original innovators of the basic theories of AI. European countries have generally laid a solid foundation of AI ethics and governance. Japan has secured outstanding achievements in the relationships between human beings and AI development. China, as one of the leading centers of AI development, should assume the responsibility for empowering the development of AI in less-advantaged countries, especially those that are still developing, to deliver the benefits of AI to every corner of the world. Other countries may draw on these insights as well.
Both Chinese and British scholars have observed that despite differences in values, principles or other unsolved divergences, cross-cultural cooperation on AI it is still effective as long as a consensus is reached on the handling of specific issues. The philosophy of seeking harmony without uniformity and providing mutual support through tough times provides communication channels for technological development, regardless of cultural and historical differences.
A global, interconnected landscape of AI development featuring multiple main centers is bound to emerge in the future. To promote this vision, countries should engage in cooperation to strengthen mutual trust, thus avoiding misunderstandings. In order to realize the global sustainable development of humankind, society, the environment and technology, we should give full expression to the positive role of highly inclusive international organizations such as the United Nations. To realize the global development of AI technology and its applications to serve the well-being of humankind and a better future for all of us, we must stand and hold together as a human community with a shared future.
Professor, Institute of Automation, Chinese Academy of Sciences.
Director of Research Center for AI Ethics and Sustainable Development, Beijing Academy of Artificial Intelligence.
Chief Scientist, Institute for AI International Governance, Tsinghua University.
Adjunct Research Fellow, Center for International Security and Strategy, Tsinghua University.
A shortened Chinese version of this opinion piece is published in Guangming Daily, Jan 20th, 2021.