Tue Jan 28 07:40:00 UTC 2025: ## Chinese Startup DeepSeek Shakes Up AI Market, Triggering Nvidia Stock Plunge and Meta’s “War Room” Response
**San Francisco/Beijing** – The AI landscape has been rocked by the emergence of DeepSeek, a Chinese startup that has unveiled a remarkably cost-effective AI model, triggering a near 17% drop in Nvidia’s stock price and wiping out nearly $600 billion in market capitalization. Nvidia, once one of the world’s two most valuable companies, saw its valuation plummet below $3 trillion following the news.
DeepSeek’s free and open-source AI model, V3, and its paid counterpart, R1, were trained using older Nvidia chips, achieving comparable performance to OpenAI’s o1 model at a fraction of the cost – a claimed 96.4% reduction. The company reports developing the model in just two months at a cost under $6 million. V3 even briefly topped the US App Store charts.
This breakthrough has sent shockwaves through the industry. Investors are questioning Nvidia’s future dominance, given DeepSeek’s ability to achieve comparable results at a significantly lower price point. Experts compare the situation to releasing an iPhone-equivalent phone for $30 instead of $1000.
The development has also prompted a swift and significant reaction from Meta. CEO Mark Zuckerberg reportedly mobilized engineers into four “war rooms” to analyze DeepSeek’s success, specifically focusing on how its backer, High-Flyer Capital, achieved such low costs and the data used to train the model. Meta aims to leverage these insights to improve its own Llama AI model. Meta confirmed they are continuously evaluating competing models.
The incident highlights concerns about the massive investments in AI by US companies, given DeepSeek’s seemingly disruptive cost-effectiveness and open-source approach. Experts like Meta’s Yann LeCun suggest this points to the growing power of open-source AI over proprietary models. Microsoft CEO Satya Nadella also noted the impressive capabilities of DeepSeek’s model at the World Economic Forum in Davos, urging a serious consideration of these developments from China.
DeepSeek’s rapid success is also attributed, in part, to US restrictions on the export of advanced AI chips to China, which inadvertently spurred domestic innovation. The company, founded in May 2023 and fully funded by High-Flyer Capital, shows no plans for further funding rounds, focusing instead on technological advancement. Its mission, notably different from other Chinese AI companies, is explicitly focused on unlocking the mysteries of Artificial General Intelligence (AGI).