As a new type of artificial intelligence technology, generative artificial intelligence has continued to gain momentum this year, attracting widespread attention from all sectors of society. On July 13, 2023, the Cyberspace Administration of China, together with seven other departments including the National Development and Reform Commission, the Ministry of Education, and the Ministry of Science and Technology, issued the "Interim Measures for the Administration of Generative Artificial Intelligence Services" (hereinafter referred to as the "Measures"). These Measures officially came into effect on August 15. This is the first time my country has made clear provisions for the research and development and services of generative artificial intelligence, and it is also the world's first legal document specifically regulating generative artificial intelligence. Compared with previous artificial intelligence technologies, generative artificial intelligence exhibits typical characteristics such as strong human-computer interaction and strong corpus dependence. It faces many problems and challenges in both technological development and risk control, urgently requiring in-depth research.
What is generative artificial intelligence?
Artificial intelligence (AI) can be categorized into decision-making AI and generative AI based on its models. Decision-making AI is a technology used for decision-making, utilizing machine learning, deep learning, and computer vision to address problems in specialized fields and help businesses and organizations optimize their decisions. Generative AI, on the other hand, is an AI technology used to generate new content. Due to breakthroughs in large models, computing power, and training methods, it can automatically generate new text, images, audio, and video content using language models, image models, and deep learning. Generative AI enables high-quality, efficient, and diverse content production, becoming a significant force driving digital productivity transformation. Currently, in terms of applications, the generative AI industry chain is gradually taking shape: its upstream includes data providers, algorithm model institutions, a creator ecosystem, and underlying supporting tools; the midstream includes processors of digital content such as text, images, audio, and video; and the downstream mainly consists of various digital content distribution platforms, consumers, and related service providers. In fields such as education, finance, healthcare, and industry, generative artificial intelligence can accelerate the integration of data and reality, promote industrial upgrading, automate the production of financial information and product introduction videos, and support generative derivative designs, thereby reducing costs and increasing efficiency. Generative artificial intelligence is playing an increasingly important role as a production tool and emerging productive force.
Generative artificial intelligence brings both tremendous opportunities and challenges.
Looking at the major technological development trends and investment hotspots both domestically and internationally in the first half of 2023, generative artificial intelligence (AI) has emerged as one of the most eye-catching technologies. Generative AI not only has a significant impact on the global economy but also brings new opportunities to various industries. Among them, the retail and consumer goods industry, the banking industry, and the pharmaceutical and medical products industry have the greatest development potential. Currently, my country has developed over 80 large-scale language models, with major participating companies including large technology companies, universities, and telecommunications equipment manufacturers. As of May this year, according to incomplete statistics, more than 40 companies and institutions in China have released generative model products or announced generative model R&D plans. The explosive growth in the number of large-scale model projects demonstrates the rapid development of generative AI services in my country. In the business field, the application of generative AI transcends organizational boundaries, deeply integrating into existing organizational workflows and systems, reconstructing key business processes in areas such as customer service, supply chain, and cybersecurity, forming influential new models and businesses. Empowered by generative AI, fields such as biopharmaceuticals, manufacturing, and materials science are beginning to change their business processes. Generative artificial intelligence has driven the process of utilizing and developing the value of massive amounts of data. In terms of information generation, editing, and iteration, it can efficiently collect and integrate various types of data, and automatically generate text content and images. It is predicted that by 2025, the proportion of synthetic information in the external marketing information of large enterprises and institutions will rise to 30%; and by 2030, the proportion of AI-generated content in major film and television works will rise to 90%.
While generative artificial intelligence (AI) is developing rapidly, it also faces various challenges. Common criticisms of its generation include inaccurate content, low feasibility, direct text compilation, and a lack of logical coherence. Furthermore, while generative AI can improve work efficiency, it also creates significant unemployment. A research report released in March of this year indicates that 300 million jobs globally are expected to be replaced by generative AI, with lawyers and administrators being the most likely to be laid off. Physically demanding or outdoor professions, such as construction and maintenance, will be less affected. Generative AI also challenges existing legal regulations. The biggest challenge in the legal field for generative AI is the potential for various infringement disputes. The generative capabilities of AI come from data and models, and the data sources themselves may infringe on the copyrights, portrait rights, reputation rights, personal information rights, and other rights of others.
Overall, generative artificial intelligence is still in its early stages, and its application prospects are unclear. We should view its development patterns objectively and create an inclusive and prudent regulatory environment for the development of new business models and technologies, so as to further stimulate market vitality and social creativity.
China's Solution for AI Governance
Data from the 2023 World Artificial Intelligence Conference shows that my country's artificial intelligence industry is booming, with the core industry reaching a scale of 500 billion yuan, more than 4,300 enterprises, and a continuous stream of innovative achievements. The development of generative artificial intelligence signifies that AI is evolving from specialized intelligence to general-purpose intelligence, entering a new stage of development. The Political Bureau of the CPC Central Committee held a meeting on April 28, clearly pointing out the need to attach importance to the development of general-purpose artificial intelligence, create an innovative ecosystem, and pay attention to risk prevention. It is believed that with the implementation of the "Measures," the development and application of generative artificial intelligence in various fields will become healthier and more standardized. Facing the opportunities and challenges brought by technological development, the "Measures," based on institutional innovation, emphasizes both information content security and promotes the development of the technology industry from multiple aspects. It not only provides a Chinese solution for international AI governance practices but also makes a positive and beneficial exploration for my country's AI legislation. The "Measures" adhere to the principle of giving equal importance to development and security, and combining the promotion of innovation with governance according to law for generative artificial intelligence. It takes effective measures to encourage the innovative development of generative artificial intelligence and implements inclusive, prudent, and classified hierarchical supervision of generative artificial intelligence services. Regarding data sources for generative artificial intelligence, the Measures stipulate the need to promote the orderly opening of public data in a classified and graded manner, expand high-quality public training data resources, and encourage the use of secure and reliable chips, software, tools, computing power, and data resources. The Measures aim to safeguard national security, public interests, and the legitimate rights and interests of citizens while using scientific and reasonable measures for supervision, thereby promoting the positive and healthy development of the artificial intelligence industry.
Although there are significant differences between domestic and international generative AI in areas such as large-scale model development, application deployment, and business models, it is believed that my country's related technologies and applications will gradually mature in the future, and the country will further explore business models suitable for its own development. The healthy development of generative AI relies on the joint efforts of the government, enterprises, society, and netizens; only in this way can this disruptive technology better benefit the people.
