China has the edge over the U.S. in mass-producing semiconductor chips and generating ways to apply AI, even though U.S. chips are slightly more effective, the Bridgewater Associates founder explained on an episode of the Tucker Carlson show last month. 

“We design chips, but we can’t produce chips effectively. By and large, we can’t produce things—any manufactured goods—as cost effectively,” Dalio said.

Instead, the U.S. advantage in AI chip development rests in its continued investment in higher education and research, which has attracted a global workforce, he argued.

“If we could work well together in that inventiveness—with rule of law working and all of that working—we have those things that are our competitive advantage,” Dalio added. “We do not have manufacturing, and we’re not going to go back and be competitive in manufacturing with China in our lifetimes, I don’t believe.”

The AI development war between the U.S. and China has been heating up. In January, Chinese dark horse developer DeepSeek unveiled an open-source model claiming to be faster and far cheaper to produce than OpenAI’s o1 model. DeepSeek’s assertions—though some experts question them—are in spite of the Biden administration’s sweeping export controls designed to restrict access to key chip manufacturing equipment and curb domestic chip production.

The U.S. has pushed to ramp up chip manufacturing at home. The Taiwan Semiconductor Manufacturing Co. (TSMC), which produces most of the world’s advanced semiconductor chips, is poised to announce a $100 billion investment in U.S. chip plants, according to multiple reports, part of President Donald Trump’s plan to push the U.S. ahead of AI competition.

Dalio called AI a “war no country can lose” in January, with winning “more important than profits,” he told the All-In podcast.

Where China has the AI edge over the U.S.

While China’s chipmaking efforts tend to result in semiconductors with small deficiencies compared to U.S. technologies, those chips are still significantly less expensive compared to their American counterparts—and China will likely shift its focus to AI applications, Dalio said.

“The Chinese play is going to be chips, very inexpensive chips embedded into manufactured goods” like robotics, Dalio said in the January podcast.

DeepSeek’s optimistic outlook has boosted the confidence of Chinese investors, including on the potential of China-based robotics companies. Jacqueline Du, Goldman Sachs’ head of China industrial technology research, told Bloomberg Television last week that robotics was the “best embodiment of artificial intelligence technology” today and said China in particular was an “excellent playground” for investors to look for opportunities.

For now, the U.S. still has the edge over China in AI research infrastructure. According to the Stanford AI Index, which measures development and technological performance among other factors, the U.S. has invested $67.2 billion in AI research compared to China’s $7.8 billion, and has surpassed China in research publications pertaining to the technology.

“The gap is actually widening” between the two countries, Ray Perrault, a computer scientist and director of the steering committee behind the index, told the Associated Press. “The U.S. is investing a lot more, at least at the level of firm creation and firm funding.”

The U.S.’s advantage in AI research brain power means China won’t be able to notch a decisive victory in the AI war, Dalio predicted. He painted a picture of the future of AI, in which there’s sustained efforts in guarding intellectual property surrounding the technology.  

“Different entities are going to be ahead in different ways,” Dalio told Carlson. “We’re going to then be in this world in which there’s competition…and then there’s an attempt to be protectionist or whatever, or to fight those differences. And that’s what the world looks like.”