Billionaire investor Ray Dalio says this will be China’s play in the AI tech wars
Bridgewater’s founder, Ray Dalio, talked about DeepSeek, the tech wars, and large-scale AI companies like Nvidia in a recent podcast interview.
The billionaire investor Ray Dalio said he expected China to leverage its manufacturing might in the AI arms race, following a similar strategy as it did with electric vehicles.
Nvidia and other large-scale AI companies have “risk issues,” the Bridgewater founder said in Tuesday’s episode of the “All-In Podcast.”
“You want to invest in productivity, but there’s great disruption, great disruption, that’s going to take place, and they’re going to be the disruptors and the disruptees,” Dalio said.
“I think the Chinese are a bit behind in the chips, but they’re ahead in the applications,” he added.
Asked about Chinese artificial intelligence startup DeepSeek, which dominated headlines over the weekend, Dalio said “the Chinese play is going to be chips, very inexpensive chips embedded into manufactured goods,” such as robotics.
“That’s how they fight wars,” the Bridgewater founder said, adding that the competition around AI is a “war no country can lose” and that winning it is “more important than profits.”
Dalio’s comments come as the Chinese AI lab DeepSeek has rocked Silicon Valley and the wider tech world with the release of its R1 model, which it said was trained at a fraction of the cost of industry-leading models like those underpinning OpenAI’s ChatGPT.
Following DeepSeek’s latest release, tech stocks took a tumble, with the American chipmaker Nvidia’s market cap plummeting by nearly $600 billion on Monday.
Ahead of Big Tech earnings, which kick off Wednesday, investors raised questions about companies’ massive investments in AI infrastructure and capital expenditures and whether apparently lower-cost models like DeepSeek’s could influence demand for Nvidia’s latest chips.
An Nvidia spokesperson previously told B-17 that DeepSeek is an “excellent AI advancement and a perfect example of Test Time Scaling” that illustrates how to leverage “widely available models and compute that is fully export control compliant.” The spokesperson added that AI inference “requires significant numbers of NVIDIA GPUs and high-performance networking.”
A spokesperson for Dalio did not respond to a request for comment.
The billionaire said many investors make the mistake of wanting to buy “good” things from “great” companies. But with the hindsight of the dot-com bubble, the share prices that AI and technology companies are trading at have “to be paid attention to,” Dalio said.
In the late 1990s, hype built up around internet companies before an eventual collapse in the early 2000s that erased trillions in market value.
“A great company that gets expensive is much worse than a bad company that’s really cheap,” Dalio said.
