DeepSeek was founded with the vision of providing open source AI models at a very low cost compared to similar models from US companies like OpenAI and Meta.
DeepSeek launched its new model, DeepSeek-R1, in January 2025, and it cost only about $5.6 million to develop, making it a free and widely available alternative.
Following the launch of DeepSeek-R1, US tech stocks saw sharp declines, with Nvidia shares down 12% in premarket trading, and shares of companies like Meta and Google falling sharply. The declines were a result of concerns about an AI bubble and fears of strong, low-cost tech competition from China.
DeepSeek-R1 outperformed models like OpenAI's GPT-4 and Meta's Llama on various tests, making it a strong contender in the AI market.
The launch caused panic in the markets, as DeepSeek became the most downloaded free app on app stores, surpassing apps like ChatGPT, raising concerns about US control over AI technology.
These developments have raised questions about the effectiveness of U.S. export restrictions on smart chips to China, suggesting that China can develop advanced technology at lower cost.
The economic impacts of DeepSeek have been swift and sweeping, resulting in significant losses in the market value of American technology companies and raising concerns about the geopolitical future of the AI industry.
