The distribution of the latest language model of the Chinese company Deepseek originated in the minds of investors doubts about the ability of American companies and further hold leadership in the field of AI, and most importantly, proved that high speed can be achieved at lower costs. President Donald Trump and the head of Openai Sam Altman praised Deepseek, but emphasized that the United States is capable of more.
Image source: Deepseek
At least, in his speech in Florida, Donald Trump said a lot in recent days about the progress of the Chinese company Deepseek, which has demonstrated a modern language model created with significantly lower hardware and material resources compared to American developments. According to the US president, such a precedent is beneficial for his country, since “we can do the same, without spending as many funds and getting the same result as we would like to hope.” At a meeting with the political leadership of China, according to Trump, he heard the confessions of the merits of American scientists. In his opinion, if the Chinese side managed to create a cheaper technology in the field of artificial intelligence, American companies will follow it.
One of the founders and the head of Openai, Sam Altman, said from the pages of the social network X that “Deepseek R1 is an impressive model, especially from the point of view of the ratio of the result and cost”. He immediately added: “We obviously can offer significantly the best models. To have a new competitor is really very enthusiastic. We will speed up the release of some models. ”
NVIDIA also could not resist commenting and on the eve also expressed admiration for the results of DeepSeek, emphasizing that such a high result was achieved in the conditions of export restrictions of the United States and a relatively small budget. “The work of Deepseek shows how new models can be created without violating export control,” the NVIDIA report said. The record drop in the capitalization of NVIDIA itself amid the successes of Deepseek did not upset representatives of the company. On the contrary, they noted that the creation of artificial intelligence with the possibility of reasoning will require significant quantities of the GPU of this brand, as well as high -speed telecommunication solutions.