The sudden rise of DeepSeek, a Chinese AI app, has sent shockwaves through the tech industry, the markets, and the once-unquestioned sense of American dominance in artificial intelligence (AI). This moment has been so impactful that venture capitalist Marc Andreessen compared it to the launch of Sputnik, the satellite that triggered the space race. DeepSeek rapidly became the most downloaded free app on the US App Store, sparking a massive sell-off in major tech stocks as concerns about America’s leadership in AI grew.
By the time US markets closed on Monday, shares of Nvidia, a leading AI chip designer, had plummeted by 17%, losing almost $600 billion in market value. This marked one of the largest drops in the history of the US stock market. The root of this disruption seems to lie in DeepSeek’s remarkably low development costs. While OpenAI, the creator of ChatGPT, spent $5 billion last year alone, DeepSeek’s developers claim to have built their model for just $5.6 million—significantly less than the resources invested by giants like Google and Anthropic.
This cost-effective approach raised eyebrows in Silicon Valley, where many were left speechless by DeepSeek’s rapid ascent. Some investors wondered if this presented a buying opportunity, while others questioned the validity of DeepSeek’s financial claims. Analyst Gene Munster expressed skepticism, questioning whether DeepSeek was receiving subsidies and whether its numbers were reliable. Despite these doubts, he described DeepSeek’s success as a “flex” by China and a “black eye for US tech.”
This shake-up comes just days after OpenAI’s Sam Altman and Oracle’s Larry Ellison joined President Donald Trump for a news conference that represented the peak of American optimism on AI. At the event, they unveiled Stargate, a joint venture promising $500 billion in private AI infrastructure investment. The US had been confident that its control over data centers and high-end chips would ensure its dominance in AI, despite China’s strength in rare-earth metals and engineering talent.
However, the emergence of DeepSeek has caused a rethinking of these assumptions. Trump called it a “wake-up call” for the American tech industry, arguing that cheaper AI is ultimately beneficial. This sudden disruption has also raised questions about the energy and resources required to develop AI. The advent of DeepSeek suggests that less capital, computing power, and energy may be necessary than previously thought.
DeepSeek’s model, developed using open-source software and existing technology, has thrown a wrench into the prevailing beliefs about what it takes to build successful AI. Some experts have pointed out that DeepSeek’s founder, Liang Wenfung, may have benefited from stockpiling essential chips over the years, potentially enabling the company to circumvent US policy restrictions on chip sales to China.
Despite the upheaval, OpenAI’s Sam Altman offered a more measured response, acknowledging DeepSeek’s impressive performance, especially considering the cost. He expressed optimism, emphasizing that competition is invigorating and that OpenAI would continue to deliver superior models.
This unexpected emergence of DeepSeek in the AI race feels reminiscent of the space race, where the US was caught off guard by the Soviet Union’s early successes. The next steps for the American tech sector will be closely watched, as DeepSeek’s rise has added significant momentum to the global AI competition.
