As Chinese startup DeepSeek is competing with Silicon Valley giants

On the day after Christmas, a small Chinese startup named DeepSeek revealed a new artificial intelligence system that could match the capabilities of state -of -the -art chatbots from companies like OpenAi and Google.

This in itself would have been a milestone. But the team behind the system, called DeepSek-V3, described an even greater step. In a research article explaining how they built technology, Deepseek engineers said they used only a fraction of highly specialized computer chips that major AI companies depend to train their systems.

These chips are at the center of a tense technological competition between the United States and China. As the US government works to keep the country’s leadership in the Global AI race, it is trying to limit the number of powerful chips, such as those manufactured by the Nvidia Silicon Valley company, which can be sold to China and other rivals .

Read more:
Dealbook: Deepseek forces a global reassessment of the technology market
What is Deepseek? Meet Ia Chinese who challenges global technology giants

But the performance of the Deepseek model raises questions about the unintentional consequences of US government commercial restrictions. The controls forced researchers in China to be creative with a wide range of tools that are freely available on the internet.

Chatbot DeepSeek answered questions, solved logic problems, and wrote its own computer programs as skillfully as any other available on the market, according to the reference tests that US AI companies have used.

And it was created economically, challenging the predominant idea that only the largest technology industry companies – all based in the United States – could afford to create the most advanced AI systems. Chinese engineers said they needed only about $ 6 million in gross computational power to build their new system. This is about 10 times less than the Meta Technology giant has been spent to develop its latest AI technology.

“The number of companies that have $ 6 million to spend is much larger than the number of companies that have $ 100 million or $ 1 billion to spend,” said Chris V. Nicholson, an investor of the risk capital company Page One Ventures, which focus on AI technologies.

Since OpenAi triggered AI boom in 2022 with the launch of ChatgPT, many experts and investors have concluded that no company could compete with market leaders without spending hundreds of millions of dollars on specialized chips.

The world’s leading AI companies train their chatbots using supercomputers that use up to 16,000 chips if no longer. Deepseek engineers, on the other hand, said they needed only about 2,000 NVIDIA specialized computer chips.

Chip restrictions in China forced Deepseek engineers to “train it more efficiently so that it could still be competitive,” said Jeffrey Ding, an assistant professor at George Washington University, who specializes in emerging technology and international relations.

Earlier this month, Biden administration issued new rules aimed at preventing China from obtaining advanced AI chips through other countries. The rules are based on several rounds of previous restrictions that prevent Chinese companies from buying or manufacturing cutting -edge computer chips. President Donald Trump has not yet indicated whether he will keep or revoke the rules.

The US government has tried to keep advanced chips outside the hands of Chinese companies due to concerns that they can be used for military purposes. In response, some companies in China stocked thousands of chips, while others obtained them from a prosperous underground market of smugglers.

Deepseek is run by a quantitative stock negotiation company called High Flyer. Until 2021, he had channeled his profits to purchase thousands of Nvidia chips, which he used to train his previous models. The company, which did not respond to requests for commentary, became known in China for recruiting newly graduated talents at top universities, offering high salaries and the ability to follow research issues that most arouse their interest.

Zihan Wang, a computer engineer who worked on a previous Deepseek model, said the company also hires people without computer science training to help technology understand and be able to generate poetry and answer questions in the noticeably difficult examination of admission University of China.

Deepseek does not manufacture consumer products, leaving its engineers focused entirely on research. This means that its technology is not limited by China’s most rigorous aspect of AI regulations, which require consumer -oriented technology to comply with government controls on information.

The main US companies continue to advance the state of art in IA. In December, OpenAi revealed a new “reasoning” system called O3 that surpasses the performance of existing technologies, although it is not yet widely available outside the company. But DeepSeek continues to show that it is not far behind. This month, it launched an impressive model of its own reasoning.

(New York Times sued OpenAi and its partner, Microsoft, accusing them of copyright violations of news-related news content. OpenAi and Microsoft denied these accusations.)

A crucial part of this fast -changing global market is an old idea: open source software. Like many other companies, Deepseek has made its latest open source AI system, which means it has shared the underlying code with other companies and researchers. This allows others to build and distribute their own products using the same technologies.

Many executives and commentators have argued that large US companies should not open their technologies as they could be used to spread misinformation or cause other serious damage. Some US legislators have explored the possibility of preventing or restricting this practice.

But others argue that if regulators suffocate the progress of open source technology in the United States, China will gain a significant advantage. If the best open source technologies come from China, they argue, US developers will build their systems on these technologies. In the long run, this could put China at the IMA Research and Development Center.

“The center of gravity of the open source community has been moving to China,” said Ion Stoica, professor of computer science at the University of California, Berkeley. “This can be a great danger to the US, because it allows China to accelerate the development of new technologies.