More than 600 million people in mainland China are now using large language models (LLMs). The Chinese government has allowed these LLMs for commercial use, using the same technology behind artificial intelligence (AI) products like ChatGPT.
Zhuang Rongwen, a director at the Cyber Administration of China (CAC) who provided the estimated numbers, said that generative AI services are “forcefully driving economic and societal growth.” CAC also acts as an internet watchdog in mainland China, according to a report by Xinhua News Agency.
Chinese authorities are approving one LLM daily on an average
At present, more than 188 LLMs are made available to the public with the Chinese government’s approval. This averages out to one new LLM approved daily for commercial use by the authorities over the past six months.
However, according to media reports and press releases by different companies, only 14 LLMs and AI-based business applications were given approval for commercial use in January 2024.
The increasing number of commercial LLM users shows the rapid adoption of generative AI in the world’s second-largest economy. This comes after 20 months of ChatGPT’s release by OpenAI on November 30, 2022, which triggered worldwide excitement.
In an interview, published today, Zhuang said that one of the primary responsibility of CAC is to “actively promote both the development and governance of generative AI.”
Zhuang noted that China has been working on the application of GenAI in a variety of sectors, including manufacturing, agriculture, education, healthcare, and medicine. This is being done in an effort to revitalize antiquated industries, promote genuine economic growth, and to level up with the United States in terms of technological advancement.
According to a draft strategy from Beijing, it is already planning to create at least 50 sets of standards for AI technology by 2026.
The suggested standards will cover a number of areas, including LLM training, AI safety, governance, industrial applications, computing systems, software, data centers, and the technical requirements and testing approaches for semiconductors.
The widespread growth of LLMs in China has been a topic of criticism
Back in July, Baidu co-founder, and CEO Robin Li Yanhong reiterated his statement that China has too many LLMs during a panel discussion at the World Artificial Intelligence Conference, which was held in Shanghai. According to a report by SCMP, Li also urged designers to create more useful apps that are based on AI.
Li said that,
“In 2023, intense competition among over 100 LLMs emerged in China, resulting in a significant waste of resources, particularly computing power.”
During the same conference, Yan Junjie, the founder and CEO of MiniMax, an AI unicorn, predicted that significant industry mergers would occur in the future. He said the LLMs would mostly be founded by around five firms only. As smaller players wil either join the bigger ones or fade out with time.
