US President Donald Trump said that after the emergence of artificial intelligence (AI) model, the rise of Chinese companies Deepseek “awakening calls” caused a shock wave of Wall Street in the US technology industry.
The stocks of major technology companies such as NVIDIA fell sharply, and the market value of chip giants lost more than $ 50 billion (£ 40 billion).
DeepSeek’s statement excited the industry, that is, its R1 model is a small part of its competitors’ cost-put forward questions about the future of American AI rule and the investment scale of American companies.
Only one week after DeepSeek was launched, Deepseek has become the most free application in the United States.
Trump responded to the news on Monday that the latest development of China’s artificial intelligence industry may be “positive” in the United States.
He told the reporter in “Air Force One”: “If you can be cheaper, if you can do it, then (and) can reach the same final result. I think this is a good thing for us.”
He also said that he was not worried about the breakthrough and added that the United States would still be the main players in the field.
On Tuesday, the stocks, including SoftBank and Tokyo Electron, have a sharp decline in stocks of Japanese AI related companies, which helped reduce the benchmark Nikkei 225 by about 1 %.
Several other markets in Asia were closed due to the New Year holidays of the lunar calendar.
The financial market in mainland China will be closed today and will be reopened on February 5.
The stock exchanges of major chip production countries in Taiwan and South Korea are also closed today.
DeepSeek provides power from the open source DeepSeek -v3 model. Its researchers claim that the training volume of the model is about $ 6 million (£ 4.2 million), which is significantly less than the billions of dollars spent by competitors. However, this proposal was raised in other people in AI.
Its emergence is because the United States restricts the sales of advanced chip technology with AI power supply.
In order to continue working without a stable import of advanced chips, Chinese artificial intelligence developers shared their work with each other and tried new technical methods.
This leads to much less computing power required by the AI model than before.
This also means that their costs are far lower than before, which may subvert the entire industry.
“Despite the access of high -end hardware, the ability of DeepSeek to compete with the American model shows that the originality and data efficiency of software can make up for hardware constraints,” said Marina ZHANG, associate professor of the University of Science and Technology of Sydney, said that he focused on China’s high -tech industry, He is China’s high -tech industry, and he is China’s high -tech industry. Tell Broadcasting Corporation
After starting DeepSeek-R1 earlier this month, the company boasted that one of the “latest models with Openai” bragged for tasks such as mathematics, coding and natural language reasoning.
DeepSeek’s technology has been praised by high -known characters.
The Chinese company claims that it can train 2,000 professional chips, while the lead model is estimated to have 16,000.
But not everyone believes. Some people doubt some of Deepseek’s claims, including Tech Mogul Elon Musk.
He responded to an article that claimed that Deepseek actually had about 50,000 NVIDIA chips. These chips have now been banned from experts to China and said: “Obviously.”
The company was founded by LIANG WENFENG in the city in southeast China Hangzhou in the southeast of China.
The 40 -year -old information and electronic engineering graduates also founded hedge funds that support DeepSeek.
I saw him at a meeting between industry experts and Chinese Prime Minister Li Qiang (Li Qiang).
In an interview in July 2024 Chinese collegeMr. Liang said he was surprised by the reaction of the previous version of the AI model.
He said: “We didn’t expect pricing would become a sensitive question.”
“We just follow our own pace, calculate costs and set prices accordingly.”
