In 2017, China observed in amazement, and surprise, like Alphago, an artificial intelligence program backed by Google, defeated a Chinese prodigy in a complex table game, go. The decisive loss against a foreign computers program, which had defeated a South Korean player, was a kind of Sputnik moment for China.
That year, Chinese officials established a bold plan to lead the world in AI for 2030, promising thousands of companies and technology -centered researchers. From this fervor arose Deep Speeek, the new Chinese company largely unknown that overturned the technological panorama creating a powerful AI model with much less money than experts had believed possible.
Deepseek is private, without apparent state support, but its success embodies the ambitions of China’s main leader, Xi Jinping, who has urged his country to “occupy the dominant heights” of technology. Mr. XI wants the Chinese economy to be driven not by old growth engines such as real estate fed with cheap debt and exports, but by the most advanced technologies such as AI, supercomputing and green energy.
For Mr. XI, this moment helps to groan the aura of superiority that the United States has had in AI, a critical field in a fierce rivalry of superpowers. China has presented itself as a benevolent global partner for developing countries, willing to share their knowledge, with Mr. XI saying that AI should not be a “game of rich countries and the rich.”
Now, Deepseek has shown that it could be possible that China will be cheaper and more accessible to everyone. However, the question is how the ruler Communist Party manages the emergence of a technology that one day could be so harmful that it could threaten their interests and control over power.
The Chinese AI regulation has varied in intensity over the years, depending on where the country evaluates its strengths and weaknesses. When the Chinese government was worried that it had stayed behind the United States in 2022 after the launch of OpenAi Chatgpt, a more hobby approach was taken that finally allowed companies such as Deepseek and others to prosper.
Now that the pendulum has turned to the other side, that confidence in the industry could be a “double -edged sword,” said Matt Sheehan, who studies China as a me Carnegie Endowment for International Peace member.
The “central instincts of the party are towards control,” said Sheehan. “As they regain confidence in China’s capabilities, they can have difficulty resisting the impulse to adopt a more practical approach to these companies.”
As to underline that possibility, the founder of Deepseek, Liang Wenfeng, was invited to a discussion with Prime Minister Li Qiang on January 20, the same day that the company launched its last and most powerful model of AI, known as R1 .
Mr. Liang’s assistance was even more noticeable considering that Depseek had not been considered one of the so -called Chinese Tigers. This distinction is reserved for high profile companies such as Zhipu AI, a new company based in Beijing that has received a substantial state investment.
Deepseek is no stranger to the impulse of the party to interfere; That may have inadvertently played a role in its eventual success. Depseek had originally trained his AI models to make bets in the Chinese stock market. But when the regulators pointed to such behavior, he turned in 2023 to advance to AI to comply with China’s industrial policy.
Then he surprised the world by rivaling the performance of its American competitors despite using much less advanced computer chips that are difficult to obtain for China, a technological feat that until recently had not been available. At home, Chinese commentators have delayed the achievement of Deepseek as evidence that American restrictions on chips exports to China are ultimately useless (despite the fact that the founder of the company has said that such limits are an important concern).
Even the recent accusations of OpenAi that Depseek incorrectly reaped their data to build their models have not deteriorate their fans in China, who accuse San Francisco’s company to spread rumors.
“The technological sanctions of the United States against China have left China with no choice but to develop,” said Sun Chenghao, an expert in foreign relations at the University of Tsinghua in Beijing, echoing a popular feeling in China. “We can only trust ourselves.”
AI occupies a special place in Mr. XI’s vision of the rise of China, with its potential to help the country overcome many of its greatest challenges such as its workforce in reduction. China has used facial recognition and algorithms to overcome its ability to examine its people and turn off the dissent. Technology is also taking into account China’s military modernization with autonomous weapons systems and even battlefield strategy.
The Deepseek development could also advance in China’s geopolitical objectives. Deepseek uses an open source model, which means that anyone can match under their hood and use their technology, unlike leading US companies that use a more expensive patented software.
“The low cost and open source nature of the Depseek model reinforces the narrative of the Chinese government that China is the place where developing countries can look for AI solutions,” said Sheehan.
How big is the player that China turns on the overall stage of AI could ultimately depend on how the government decides equilibrium regulations With the freedom that companies and researchers need to do a avant -garde job that allows them to compete with the United States.
Some analysts such as Gregory C. Allen, a researcher at the Center for Strategic and International Studies and former United States defense official, said there were probably no restrictions on the development of AI when it comes to China’s army.
“The only thing that stops them is performance,” said Allen, who in his former work held conversations with members of the popular liberation army responsible for evaluating the risks of the AI
The same is not true to regulate AI in the private sector. The panorama is dictated by the competitive priorities of China’s regulatory agencies, each feeling a technology that many in the world still do not understand completely.
It is clear that the more used a technology is, the more the party will want. Users who require “socialist fundamental values” and avoid the information that undermines “state power.”
In the case of Deepseek chatbot, this has led awkward answers to seemingly benign questions such as: “Who is Xi Jinping?” Researchers who prove their capabilities have found that the bot offer answers that disseminate Chinese propaganda and even parrot misinformation campaigns.
Some concerns are more existential in nature. A growing choir of academics has been playing alarm on the potentially catastrophic consequences of losing human control over AI
The main of these voices has been Andrew Yao, a giant in AI at the University of Tsinghua and a recipient of the Turing Prize, the equivalent of the Nobel Computer Prize. His influence helped establish what China calls the global Government initiative of AI, which was introduced by Mr. XI in 2023 and included a call to always maintain AI under human control. Last year, the Government also requested the improvement of the governance of AI “based on decision -making and human supervision.”
Ultimately, AI in China can only advance to where the government decides that it can mitigate these risks, said Barath Harithhas, an AI policy expert at the Center for Strategic and International Studies, a group of Washington experts.
“The envelope and the need to adhere to the ‘central socialist values’ could risk the potential of the AI,” said Harithas.
