Wall Street saw a major upheaval when the US stock market saw record-breaking losses, which were attributed to DeepSeek, a Chinese AI business. The company’s AI developments are upending the IT industry and igniting heated discussions over the geopolitical and economic rivalry between China and the United States in AI research.
DeepSeek: What is it?
DeepSeek was established in Hangzhou, China, in 2023, and gained notoriety that same year thanks to the introduction of its vast language model. Liang Wenfeng, the CEO of the company, was a co-founder of High-Flyer, a top hedge fund in China that specializes in AI-driven quantitative trading. In order to create and operate AI systems, High-Flyer accumulated 10,000 of NVIDIA’s high-performance A100 graphics processor processors by 2022. However, sales of these chips to China were immediately limited by U.S. export restrictions.
DeepSeek adjusted by utilizing NVIDIA’s H800 chips, which are nevertheless accessible in China, in spite of these limitations. This tactical change showed that state-of-the-art hardware is not always necessary for cutting-edge AI research.
The Effect on the Market: Trillions Lost
When DeepSeek’s inventions spread throughout the stock market, it resulted in a historic collapse. The stocks of major tech companies fell:
- NVIDIA: down 16.47%
- Broadcom: 16.18% lower
- Oracle: down 10.95%
- Microsoft is down 3.71%.
- Google: down 2.76%
These losses highlight the disruptive impact of DeepSeek’s AI innovations, raising concerns among experts and investors about whether a new era ruled by Chinese people is about to begin.
The DeepSeek R1 Model is the catalyst.
Last month, DeepSeek made headlines around the world when they unveiled a new AI model that they said could compete with those from American firms like OpenAI. The model’s efficiency and scalability were demonstrated by its economical utilization of NVIDIA hardware. But the thing that really rocked the business was the publication of a follow-up study on the R1 model.
The R1 model demonstrated sophisticated reasoning abilities, including reevaluating methods for solving challenging mathematical issues. The U.S. tech industry’s expenditure tactics were called into question by its capabilities and notably reduced prices when compared to OpenAI’s models. According to Stacy Rasgon, a Bernstein analyst for the semiconductor sector, “the price points freaked people off.”
Geopolitical Consequences: The Sputnik Moment of AI
The fervor around DeepSeek has been compared to the Cold War space race that began with the 1957 launch of Sputnik. According to U.S. President Donald Trump advisor and venture capitalist Marc Andreessen, the R1 model AI’s “Sputnik moment”
According to Andreessen and others, the United States’ overregulation of AI could stifle innovation and give China the advantage. Gregory Allen, head of the Center for Strategic and International Studies’ Wadhwani AI Center, pointed out that DeepSeek’s announcement had political overtones and likened it to Huawei’s calculated actions during the trade spat between the United States and China.
Allen remarked, “The timing of the release is political in character,” implying that China wants to show how pointless American export restrictions are.
Trump’s Reaction: A Warning
Speaking to House Republicans in Miami, President Trump called DeepSeek’s advancements a “wakeup call” for American businesses. He emphasized the necessity of making strategic investments in AI development, saying, “We need to be laser-focused on competing to win.”
In keeping with his administration’s tough stance on tech competition with China, Trump also signed an executive order to close export control loopholes.
What Makes DeepSeek Unique?
In contrast to its American rivals, DeepSeek’s models are open-source, meaning that the general public can access and alter important elements of them. Although this transparency has garnered praise, it has also sparked worries about possible technological abuse.
The R1 model has received special recognition for its capacity to “think out loud” and improve its training without the need for further data sources. According to Rand Corp. researcher Lennart Heim, it is a “great example of Test Time Scaling.”
Even while American firms like OpenAI and Google have comparable capabilities, DeepSeek’s quick development highlights how the gap between Chinese and American AI research is closing. “I thought OpenAI was the front-runner,” Heim remarked. “It turns out that this is not entirely true.”
The Path Ahead
DeepSeek’s innovations have vast ramifications that impact not just the technology sector but also more general geopolitical tactics. The stakes are bigger than ever as China and the United States compete for supremacy in AI.
The issue still stands: Will China assume the lead in AI innovation in the coming years, or can the United States hold its advantage?