During a visit to the United Nations in September, Chinese Foreign Minister Wang Yi launched a new international plan to develop AI focused on what has become known as the “Global South.” Wang Yi announced China’s determination to cooperate with other countries on data sharing and security, AI technology production and supply, building joint AI research institutes, educational exchanges, and building an “open source AI community”. .
The announcement of the “Artificial Intelligence Capability Building Program” came on the heels of a UN resolution put forward by China and a “core group” of partner countries to strengthen international cooperation on AI. It positioned China as an alternative to the current industry leader, the United States.
In his speech about the plan, Wang emphasized that AI should be an inclusive and accessible technology. This was not a very sophisticated attack on the United States. Restrictions on U.S. high-tech exports have been frequently featured in Chinese state media as a means not only to target China but also to assert U.S. superiority over all other countries. “AI should not be a tool to maintain hegemony or seek superiority,” Wang said.
However, China’s actions suggest that it may have grand, if not hegemonic, ambitions of its own. The state-run People’s Daily boasted that Wang’s AI plan “demonstrates China’s vision and responsibility as an AI superpower.” Government rhetoric draws a direct line between AI exports and existing efforts to expand China’s influence abroad, such as Xi Jinping’s signature Belt and Road Initiative (BRI) and Global Development Initiative (GDI). doing. In this case, the more influence China has over AI abroad, the more likely it is to influence technological development in other countries.
open access development
Wang Yi’s speech, like other speeches on AI by Chinese representatives at the United Nations, reflected Xi Jinping’s instructions on how the technology should be applied globally. In October, Mr. Xi launched the “Global AI Governance Initiative” at the Belt and Road High-Level Conference. In his speech, Mr. Xi described the initiative as an achievement of the Belt and Road Initiative and another means of creating what he called a “community with a shared future for mankind.” An accompanying policy document outlines how AI should be “human-oriented” ( い人人で本) and a form of “intelligence for the good” ( 智亚圖子). In other words, AI should be used to benefit humanity. It must offer solutions to global problems such as social inequality and climate change, and it must be kept out of the hands of malicious actors.
The document also states that AI must not be used to interfere in the internal affairs of other countries. This is a term that China has invoked for as long as it has existed to enlist countries in the Global South in its ongoing efforts to seize Taiwan. This is to fend off international criticism of its human rights record. The document also emphasizes that all countries should be able to harness the power of AI for their own development. Therefore, AI should be open access and free for all, and all countries should have equal opportunity to access the technology.
As China’s special envoy to the United Nations, Hucong, said in his opening speech to the Friends of the World in November, AI must not become “a game for rich countries and wealthy people.” This is not just a question of the United States versus China, or the West versus the rest of the world.
China’s decision to jointly launch an AI capacity building program with Zambia also had a symbolic element. The African country received thousands of Chinese workers and hundreds of millions of RMB in loans in the 1960s, making it the beneficiary of China’s earliest overseas infrastructure projects, according to Chinese state media reports. . This is another thread connecting modern AI cooperation with China. China has long-standing ambitions to lead the developing world while becoming a superpower in its own right. In a 2018 meeting with Zambia’s president, Xi said the two countries needed to “protect the common interests of developing countries.”
AI allies
There are several possibilities for what the latest iteration of cooperation between China and the Global South could look like in terms of AI. One method is to use large-scale language models (LLMs). China is currently leading the way in open source foundational LLMs, AI models that software developers can copy and modify for free and use as the basis for building their own apps. Users report that Chinese LLMs are not only cheaper than their Western counterparts, but also better at handling East Asian languages and coding.
China’s high-tech industry leaders know that having a monopoly on this key software will put them in pole position in the future. Li Kaifu, founder of Chinese AI company 01.AI, said at a recent industry summit that China can lead from a strong position by mastering “pre-training of large-scale models.” As long as China’s models are “good enough and cheap enough,” China will be in a position to determine the “finish line of safety and control.” That could mean they have to follow the Chinese Communist Party’s political red lines no matter where they are in the world.
In a sense, this process has already begun. Egypt, one of the core members of the Group of Friends, has been expanding its AI program since at least 2019, collaborating with various foreign companies to develop AI capabilities. In May, Huawei launched public cloud services for the country and the world’s first Arabic-specific LLM. Huawei says it uses the “world’s largest” Arabic dataset. In September 2022, Saudi Arabia also signed a joint venture agreement with Chinese AI company SenseTime to establish a localized AI lab.
As the strategic conflict between the United States and China draws much of the world into competing camps, China is simply “seeking advantage,” as Wang Yi said in a lofty speech at the United Nations. Instead, AI has become one of the key battlegrounds. , but you’ve already found it.
