Donald Trump has praised China’s new AI breakthrough as a “wake-up call” for American tech firms that ultimately “could be good” for the U.S.
Shares in hardware maker Nvidia and numerous other Silicon Valley companies plunged by a total of more than $1 trillion after Hangzhou-based start-up DeepSeek released its latest AI model.
Known as R1, the model reportedly beats rivals such as Meta and ChatGPT creator OpenAI in several key metrics while costing just a fraction to train and develop.
While he issued a warning in a speech at a Republican retreat in Miami on Monday, President Trump was also bullish about American companies’ ability to take the pressure.
“The release of DeepSeek AI from a Chinese company should be a wake-up call for our industries — that we need to be laser-focused on competing to win,” he said.
“We have the greatest scientists in the world. Even Chinese leadership told me that …. ‘Silicon Valley,’ they said, ‘there’s nobody like those people.'”
He then said the breakthrough could be “very much a positive development” because it would save American companies “billions and billions” of dollars, though it wasn’t clear exactly what he meant.
“Under the Trump administration we’re going to unleash our tech companies and we’re going to dominate the future like never before,” he vowed.
The news comes after Trump scrapped Joe Biden’s 2023 safeguards on AI development and announced a new $500 billion investment fund for new data centers and power plants to fuel the industry.
He has previously argued that although it’s possible AI will “take over the human race,” American companies must “take the lead over China” and dominate the technology.
Like ChatGPT, DeepSeek R1 is a “large language” model — a type of “generative” AI that uses neural networks to digest and process vast quantities of human-generated text and learn how to create its own.
Unlike ChatGPT, though, R1 is open-source, meaning that its code can be freely studied by anyone. That stands in contrast to OpenAI and Anthropic, whose executives have argued that open-source AI models are potentially more unsafe for the world.
Over the weekend DeepSeek’s chatbot app shot to the top of Apple’s App Store charts, overtaking ChatGPT as the top-rated free iPhone application in the US.
“DeepSeek R1 is AI’s Sputnik moment,” wrote venture capitalist Marc Andreessen on X on Sunday, referring to the Soviet Union’s shock satellite launch in 1957 that galvanized the space race.
One anonymous worker at Meta, the parent company of Facebook and Instagram, claimed on an insiders’ forum that Meta was in “panic” mode and was “moving frantically” to “copy” everything it could from the new rival.
U.S. tech firms have had to dig deeply into their pockets to fund the vast expense of training more powerful generative AI systems and running existing ones, buying or renting ever-growing armadas of specialized microchips to perform the energy-intensive work.
Heightening the tension is the fact that generative AI has so far largely failed to turn a profit, leading some analysts to believe that the current hype is actually a bubble.
American tech bosses put on a brave face on Monday. OpenAI boss Sam Altman called R1 “an impressive model, particularly around … price,” but promised to deliver “better” alternatives and said it was “invigorating to have a new competitor.” A spokesperson for Nvidia likewise called R1 “an excellent AI advancement.”
Meanwhile, Chinese social media users expressed jubilation at DeepSeek’s success, according to BBC News, hailing it as “the real deal” and “the best new year gift” for the country.