Too late now for U.S. to hold back China in global AI race
Government and public embrace smart tech for efficiency and reduced corruption
Zhou Xin is senior vice president for public affairs at artificial intelligence-driven molecular imaging company Evomics Medical in Shanghai and executive chief editor of AI community platform The Yuan.
Believing that artificial intelligence could make China a greater strategic threat, the administration of U.S. President Joe Biden has set its sights on holding back its rival's AI development.
Recent measures have included restricting U.S. investors from putting money into Chinese AI companies, downgrading scholarly exchanges and cutting off supplies of specialized computer chips. Officials are also discussing limiting access by Chinese companies to U.S. cloud computing services that might provide indirect access to AI technologies.
Washington is working to extend its "great firewall" against the Chinese AI threat to allies such as Japan, South Korea and Taiwan, which previously focused more on doing business with Beijing than on geopolitics. Indeed, at their three-way summit with Biden at Camp David in August, Japanese Prime Minister Fumio Kishida and South Korean President Yoon Suk Yeol affirmed they would stand with Washington.
Can Washington really stop Chinese AI development? This is highly doubtful. China created its industrial AI policy seven years ago, long before the U.S. government came up with any kind of AI overview.
While the ChatGPT platform from U.S. company OpenAI has been a focus of global attention regarding generative AI over the past year, Chinese software developers have been hard at work too.
In August, Beijing authorized the public release of new generative AI services from Tencent, Baidu, Huawei Technologies, Alibaba Group, JD.com, ByteDance, iFlytek and Kuaishou Technology. Overnight, there are now more major technology companies in China offering their own advanced chatbots than in the U.S.
More importantly, Beijing is already engaged in embedding AI throughout China's social infrastructure. Chatbots now generate calls from public service centers. Public parking lots are managed by smart systems without a need for human attendants. Hospital and other public facilities have their own AI systems to deal with the public.
Until now, China's bureaucracy was infamous for its inefficiency. To cope with the burden of serving what was until this year the world's largest population, China's leaders since ancient times have relied on a pyramidlike model of governance that implicitly tolerated a certain degree of official corruption.
When officials recognized that dilatory handling of public requests might elicit offers of bribes, they often just became even greedier and slower about their work. While the public could, to some extent, put up with greasing palms to get problems addressed more quickly, they would eventually reach a breaking point and rise up against their corrupt overseers.
AI has not just increased the efficiency of public services, but it has also put officials under technological supervision, undermining their leeway to seek bribes. This is already proving to be a better check against corruption than long-standing Communist Party calls for integrity or centuries of Confucian precepts.
Many Chinese are quite happy with swapping the old system of corruption-plagued public services for AI-driven alternatives, even if the new versions rely on the use of facial recognition scans and other biometric systems that would undoubtedly spur privacy complaints in Western nations.
The U.S. will clearly not take the same path nor embrace dictatorship even if it would bring greater efficiency. But Chinese President Xi Jinping is attuned to what his country's citizens value, in part due to his diverse experiences growing up, first as a child of elite privilege and then as a rural peasant when his family was exiled from Beijing amid the Cultural Revolution. Xi does not understand machine learning or the technology behind AI, to be sure, but has an in-depth understanding of what Chinese people hate and need.
For the American people, the equation is different. If Biden tried to adopt AI as a governance tool, that would likely cripple the technology's effectiveness. Systems like ChatGPT would come under great strain if pulled into America's bureaucracy. Labor unions, legislators, gay rights activists, racists, anti-racists and other special interests will never trust AI systems to protect the values they hold dear.
But no matter how tight U.S. export controls get in the future, AI is already taking over in China. It is transforming Asia's largest economy despite the Biden administration's best efforts. If the U.S. does not want to fall behind China in the AI race, it is America that will have to change.