According to Business Today, the announcement of V3 and R1 by DeepSeek has been described as a “Sputnik moment” in the AI industry, similar to the launch of the first Soviet satellite in 1957, raising concerns about the West falling behind. DeepSeek’s AI models show that China is not only capable of competing with, but also surpassing, the long-standing AI giants.
One of the things that sets DeepSeek apart is its low cost and high efficiency. R1 is trained on Nvidia's H800 chip, which is 96.4% cheaper than OpenAI's o1 model, but still achieves similar performance.
Notably, the Chinese company spent just two months and less than $6 million to develop ChatGPT, while OpenAI spent billions of dollars on ChatGPT. According to experts, R1 has operating costs 20 to 50 times lower than OpenAI's ChatGPT, challenging the notion that advanced AI requires the most powerful chips. In addition, DeepSeek also applies open source code, making it easy for other developers to access and create AI applications at low cost.
Yann LeCun, head of AI research at Meta, emphasized that AI DeepSeek's success is due to open research, which not only benefits China but also promotes the overall development of global AI.
The DeepSeek AI explosion has shaken the market, causing Nvidia to lose nearly $600 billion in value in a single day. Companies that have invested billions of dollars in AI development are also starting to question the real performance of the expensive AI models that are costing them money.
Uday Kotak, founder of Kotak Bank, said DeepSeek demonstrates China’s willingness to join the technology race and challenge Western dominance. Just as Sputnik transformed the space industry, DeepSeek could usher in a new era of AI, where small but innovative companies can compete with tech giants.
