On January 20, 2025, China’s DeepSeek R1 machine learning model upset stock prices by a trillion dollars and revolutionized the AI industry when it hit the public market. The model, competing with OpenAI’s ChatGPT and Meta’s Llama, was created on a budget of seven million. On the other hand, the massively popular competitors used hundreds of billions of dollars to achieve the same results. The innovation took the world by storm, causing many to speculate on whether the United States is ahead of the game on AI.
The first and most apparent impact of DeepSeek’s release was the drastic drop in tech stock prices. Nvidia, the largest producer of computer chips, lost an unimaginable 600 billion dollars in stock prices in one day. That is the largest crash for a single company in stock market history. Yet this did not only affect Nvidia. Tech companies as a whole lost just over one trillion dollars in one day. But why did the entrance of a small AI model devastate competitors? Simply, DeepSeek does more for less. Nvidia produces high-end computer chips—an integral part of AI development—yet DeepSeek uses less powerful chips. OpenAI spends billions of dollars to train and develop ChatGPT, yet DeepSeek only spent seven million. Meta uses terabytes of data in massive data centers to run Llama, yet Deepseek can run on a Macbook. The Chinese model uses the massive ChatGPT and Llama as a framework to guide its training. The knowledge from the preexisting AI giants was compressed into a tiny model that can operate on almost any device. Retired Microsoft engineer Dave Plumber compares, “[It is] like assembling a panel of experts to train one exceptionally bright student.” The efficiency of a powerful AI model built on underwhelming parts undermines the importance of Nvidia’s chips. It was assumed that machine learning companies needed to spend grand sums of money on these expensive components. However, DeepSeek’s release opened the eyes of investors to Nvidia’s massively overinflated value in AI, causing many to sell all at once. The result: a 600 billion dollar deficit in stock prices. Many other companies also felt the wave of doubt sweep over them as many semiconductor producers lost 6-17 percent of their stock prices.
Although DeepSeek was detrimental to American tech companies, Nvidia made a statement complementing the ingenuity behind its creation. It opens up a whole new avenue in the AI industry that can be expanded upon by companies with more resources.