What is DeepSeek? This new artificial intelligence developed in China has shaken global stock markets in just one day, destabilizing American tech giants and nuclear energy companies. Let’s try to understand the reasons behind this technological revolution with two illuminating metaphors.
DeepSeek marks a turning point. China has demonstrated its ability to compete on equal terms in the field of AI, a domain increasingly considered strategic for the future.
The second metaphor is the “bulldozer effect.” Imagine a hill on which American tech giants stood comfortably, protected by their massive investments and the assurance of an unchallenged dominance. In just one day, DeepSeek leveled that playing field. Not because it is more advanced than Western AIs, but because it achieves the same efficiency at a fraction of the cost.
While OpenAI and other American companies spend billions to train their language models, DeepSeek’s developers managed to do so with just $5.5 million. This is the figure that has shocked the sector: a scientific paper published by the developers reveals that their model is open source, freely downloadable, and usable by anyone.
But there’s more: DeepSeek is extremely economical to operate. For the same workload, where ChatGPT incurs a cost of $270, DeepSeek requires only $10. This remarkable efficiency has opened two cracks in the global landscape:
- The end of American technological superiority. The notion that the United States dominates the field of artificial intelligence is now being questioned.
- The economic model of American big tech. These companies continue to demand enormous funding to develop new models, claiming it is the only way to innovate. Just a few days ago, former President Trump announced a $500 billion Stargate plan.
Behind this revolution is Liang Wenfeng, a young, rebellious, and visionary Chinese developer. Liang not only created a groundbreaking model with limited resources but also found a way to update its libraries directly on competing platforms like ChatGPT, using APIs to continuously improve DeepSeek.
The ripple effects of DeepSeek have hit Wall Street hard, causing massive losses. Nvidia, the giant in graphics cards essential for AI, lost $600 billion in a single trading session. Why? DeepSeek requires far less computing power than traditional models, challenging Nvidia’s dominance in the hardware market for artificial intelligence.
It’s not just Nvidia: companies involved in nuclear energy production also saw their shares drop by 30% in a single day. The technological infrastructures of American big tech are notoriously energy-intensive, whereas DeepSeek consumes much less energy. This shift could revolutionize the balance of supply and demand in the global energy market.
However, a critical question remains: some sources suggest that DeepSeek may have used next-generation graphics cards, which theoretically should not be available in China due to export restrictions. If confirmed, this detail would raise even more profound questions about the development dynamics of this system. Nonetheless, all research material is open source, making it easily accessible for verification and further analysis.
DeepSeek has proven that innovation is not just about power but also about efficiency. With reduced costs and comparable performance, this new Chinese model has rewritten the rules of the game in the field of artificial intelligence, opening a new chapter in the competition between East and West.
The West trembles, and the future is yet to be written.
Alessandro Sicuro Comunication
for
Sure-com America
Sure-com America 