The Biden administration introduced novel export restrictions this week aimed at controlling the progress of artificial intelligence around the world and ultimately preventing the most advanced artificial intelligence from falling into China’s hands. The rule is the latest in a series of measures introduced by Donald Trump and Joe Biden to control Chinese artificial intelligence.
With prominent AI figures including OpenAI Sam Altman and anthropic Dariusz Amodei by warning of the need to “beat China” in artificial intelligence, the Trump administration could lead to further escalation.
Paweł Triolo is a partner at DGA Group, a global consulting firm, a member of the Council on Foreign Relations, and senior advisor to the Penn Project on the Future of U.S.-China Relations at the University of Pennsylvania. Alvin Graylin is an entrepreneur who previously operated in China for the Taiwanese electronics company HPC. Together, they monitor China’s artificial intelligence industry and the impact of US sanctions. In an email exchange, Triolo and Graylin discussed the latest sanctions, Silicon Valley rhetoric and the dangers of viewing global AI as a zero-sum game.
This interview has been edited for clarity and brevity.
What do you think about the novel AI proliferation rules introduced this week by the U.S. government that aim to limit China’s access to AI?
Paul Triolo: Generally focused on high performance computing clusters. This rule also places controls on proprietary model weights for the most advanced “front-end” models, but it is unclear how performance levels will be determined, and most open-weight models [freely shared] AI models are fine-tuned and improved by users, including the largest artificial intelligence companies in China.
Complicated rules and unclear terms of compliance introduce significant uncertainty into the long-term plans of both medium and immense hyperscalers in the United States and the West.
For hyperscale companies like Google, Microsoft, AWS, and Oracle, this rule introduces critical issues, including slower or more elaborate international expansion, novel compliance and legal costs, impacts on global R&D, and uncertain enforcement requirements.
How have previous measures, including sanctions introduced by the first Trump administration, affected the country’s AI industry?
Paul Triolo: US export controls have slowed China down, but at a high level sanctions have united the will and efforts of the Chinese government to become more self-sufficient. It has committed tens of billions to helping local players catch up technologically or become more proficient in key areas, resulting in significant changes to the semiconductor industry and its ability to support advanced hardware to develop pioneering artificial intelligence models.
Chinese AI developers are very good at leveraging older AI equipment from Western companies and gradually incorporating domestic alternatives into their development process. Chinese companies will continue to innovate in AI hardware and software, if not at the pace of their Western counterparts.
Why do you think so many people in Silicon Valley are now talking about the need to “beat China” in AI?
Paul Triolo: There is a growing connection between conservative venture capitalists, mainly located in Silicon Valley, and technology companies whose business models are based on advertising the threat from China. It’s a troubling combination that combines the threat from China, personal gain, and opposition to regulating advanced artificial intelligence. It also portrays the U.S.-China competition around artificial intelligence as a zero-sum game, which is particularly threatening.