Why ‘Beating China’ in AI Brings Its Risks

The Biden administration this week implemented new restrictions on exports that were created facilitating the advancement of AI worldwide and ultimately prevent advanced AI from falling into China’s hands. The order is the latest in a series of moves by Donald Trump and Joe Biden to control China’s AI.

They are popular AI statistics including OpenAI’s Sam Altman and Anthropic Darius Amodei warning of the need to “beat China” in AI, the Trump administration could escalate matters.

Paul Triolo is a partner at DGA Group, a global consulting firm, a member of the Institute of Foreign Relations, and a senior advisor to the University of Pennsylvania’s Penn Project on the Future of US-China Relations. Alvin Graylin is an entrepreneur who previously ran the Chinese operations of Taiwanese electronics company HPC. Together they have been targeting Chinese AI companies and the consequences of US sanctions. In an email exchange, Triolo and Graylin discussed recent sanctions, claims about Silicon Valley, and the dangers of viewing global AI as a zero-sum game.

This interview has been edited for clarity and brevity.

What are you making new? The law of diffusion of AI from the US government this week, which wants to limit China’s access to AI?

Paul Triolo: Generally, it’s focused on supercomputers. The law also places controls on the weights of the highest quality “borderline” models but it is unclear how performance will be determined, and many open (freely shared) AI models are modified and controlled by users, including large AI companies. in China.

The complicated law and unclear compliance conditions raise serious doubts in the long-term plans of the central and senior US and Western hyperscalers.

For hyperscalers such as Google, Microsoft, AWS, and Oracle, this law creates challenges, including slow or difficult global expansion, new compliance and litigation costs, the impact of global R&D, and uncertain enforcement requirements.

How have previous measures, including sanctions introduced by the first Trump administration, affected the AI ​​industry there?

Paul Triolo: US foreign policy has slowed China down, but at a higher level the sanctions have aligned the Chinese government’s desire and efforts to become more self-reliant. It has plowed billions into helping local players capture technology or grow in startups, which has led to significant changes in the electronics industry and its ability to support advanced tools to create frontier AI models.

Chinese AI developers have found great success by adopting AI tools from Western companies and slowly integrating some domestic techniques into their own production. Chinese companies will continue to develop AI tools and software, if not to the same extent as their Western counterparts.

Why do you think many in Silicon Valley are now talking about the need to “beat China” in AI?

Paul Triolo: There is a growing link between venture capitalists, who live in Silicon Valley, and technology companies whose businesses are based on the threat of China. This is a challenging combination that combines China’s threat, your own gain, and backlash against the control of advanced AI. It also shows the US China competition around AI as zero sum, which is very dangerous.