DeepSeek's Disruption: How a Chinese AI Startup is Shaking Up the Tech World

Imagine a single AI breakthrough causing a trillion-dollar upheaval in global markets. This is precisely what happened when China's DeepSeek unveiled its latest AI model.

The Emergence of DeepSeek

Founded by Liang Wenfeng, DeepSeek recently introduced the R1 reasoning model, which has been hailed as a "Sputnik moment" for American AI. The model's performance in tasks like mathematics, coding, and natural language reasoning is reportedly on par with leading U.S. AI models. [source]

Efficiency Redefined

What sets DeepSeek apart is its resource efficiency. While leading AI companies often train their models using supercomputers with over 16,000 GPUs, DeepSeek achieved comparable results using approximately 2,000 GPUs. This approach not only reduced training time to about 55 days but also cut costs to around $5.58 million—roughly one-tenth of what some U.S. tech giants have spent. [source]

Market Turmoil

The unveiling of DeepSeek's R1 model led to a significant selloff in global technology stocks. Companies like Nvidia, Microsoft, and Alphabet experienced notable declines in their stock prices, resulting in nearly $1 trillion being wiped off American stocks by January 28, 2025. [source]

Global Reactions

Industry leaders have had mixed responses to DeepSeek's advancements. Microsoft CEO Satya Nadella and OpenAI CEO Sam Altman described DeepSeek's achievements as "super impressive." In contrast, figures like Elon Musk expressed skepticism regarding the sustainability of DeepSeek's success. [source]