On July 9, at the World Artificial Intelligence Conference's Scientific and Technological Innovation Plenary Session, the paper "A New Generation of Artificial Intelligence in China" (referred to as "AI in China") written by a Chinese team was officially released to the public.

This paper is the first time that Chinese scientists have published a panoramic review of China's AI development in a top international academic journal. The publication of the paper reflects the global importance and influence of the development results of China's artificial intelligence industry, and also demonstrates Shanghai's strength as a global artificial intelligence highland.

In addition to mentioning the application of AI in various industries and talent training, the paper also introduces the ethical issues of artificial intelligence in detail.

The paper writes that according to China's new generation of artificial intelligence development plan, to achieve the long-term goal of artificial intelligence benefiting human society, we must not only increase the research and development and application of artificial intelligence and maximize the potential of artificial intelligence, but also pay more attention to the many ethical, legal and economic issues brought about by the development of artificial intelligence.

In June 2019, the National New Generation Artificial Intelligence Governance Committee issued the "Principles for Responsible Artificial Intelligence Governance for the Development of New Generation Artificial Intelligence." In April 2019, the European Commission’s High-Level Expert Group on Artificial Intelligence proposed ethical guidelines for trustworthy artificial intelligence. The Beijing Institute of Artificial Intelligence also published the "Beijing Artificial Intelligence Principles" in May 2019. The G20 supports the principle of responsible management of trustworthy artificial intelligence.

Speaking of the differences in artificial intelligence ethics at home and abroad, the paper states that the difference in artificial intelligence ethics between China and Europe may be that China emphasizes convenience and trustworthiness, while Europe focuses on human-centeredness and privacy. Since the development of artificial intelligence can be used for harm, China's artificial intelligence ethics are required to help build trustworthy artificial intelligence systems to facilitate daily life, which should be executed exactly as planned to ensure the safety and reliability of artificial intelligence systems. Because the development of AI relies on data to train the underlying systems, AI ethics in Europe have strong data protections, which limit the extent to which companies operating there can exploit people's personal data.

The paper states that when we talk about artificial intelligence ethics, people generally think that artificial intelligence algorithms are safe, reliable, responsible, explainable and fair. The paper's authors now realize they need to incorporate the aforementioned ethics into their research and understand the biases and prospects for social harm from algorithmic injustice. For example, research can leverage federated learning and blockchain to alleviate data privacy issues, while debiasing algorithms can ensure fairness to a certain extent. Private aggregation of teacher sets is used to enable private learning by carefully coordinating the activities of several different machine learning models.

So far, China’s artificial intelligence ethical principles lack legal enforcement and policy implications. In order to formulate ethics and policies related to the development of artificial intelligence in China, the Chinese government is determined to evaluate the impact of artificial intelligence technology on society. According to the "Work Guidelines on Establishing a National New Generation Artificial Intelligence Innovation and Development Experimental Zone" issued by the Ministry of Science and Technology of China, China will carry out long-term, interdisciplinary social experiments and explore appropriate methods and means to support artificial intelligence governance.

At the end of the paper, the authors made an outlook, hoping to cultivate an artificial intelligence ecosystem in the future. The paper states that artificial intelligence has the potential to enhance other technologies, similar to enabling technologies such as internal combustion engines and electricity. Many people believe that artificial intelligence is universal and has many applications in different fields and brings seismic changes in these fields. When we talk about artificial intelligence applications, there are many different components that support certain applications, such as software, hardware, platform, and marketing.

The artificial intelligence ecosystem can be a system-interconnected network composed of various organic parts, including start-ups, enterprises, investors, research institutions and markets, which work together to create innovative products and competitive services in an environment of collaboration and knowledge transfer. In a healthy AI ecosystem, every player across multiple industries and domains can find ways to thrive.