The tech war between China and the US, which began at the dawn of the new year, has given a run to Nvidia Corporation, a US-based advanced semiconductor manufacturer specializing in artificial intelligence (AI), and it seems that China won the first round by rolling out a cutting-edge AI technology, DeepSeek-V3 models.

DeepSeek's AI assistant is now accessible for download in the US, positioning itself as a competitor to the US OpenAI's ChatGPT and Google’s Gemini, both built LLM--an advanced AI system designed to understand and generate human-like text, trained on vast amounts of text data, and that performed various language-related tasks, such as translation, summarization, and conversation.

An AI-advanced technology—as DeepSeek is—acquires knowledge and addresses challenges, generally through extensive training on substantial datasets and the identification of patterns. The outcome is software capable of engaging in conversations akin to those of a human being or forecasting individuals' purchasing behaviors. In recent years, it has gained prominence as the underlying technology for chatbots such as ChatGPT and DeepSeek, commonly referred to as generative artificial intelligence.

At the beginning of February, DeepSeek's application ascended to the pinnacle of Apple's App Store, displacing OpenAI's ChatGPT from its leading position and maintaining its lead in the AI application stores till now.

Liang Wenfeng, the founder of DeepSeek, in an interview, told China Talk last year that as the economy develops, China should gradually become a contributor instead of freeriding. “Over the past three decades of the information technology revolution, China has largely refrained from engaging in genuine technological innovation. It’s just because we were not previously involved in this process and have disregarded its existence,” he replied when asked about the belief that the US excels at technological innovation while China excels at making applications.

Technology and stock markets shock

The stunning entry of two large language models (LLMs), DeepSeek RI, in late January by a small and unknown Chinese AI technology firm, DeepSeek, has caused significant damage to tech companies, the stock market, and the value of digital currencies.

DeepSeek has caused a significant decline in the market capitalization of AI companies, with the total value falling by more than $2.5 billion in just 24 hours of its release. This fall coincides with the rise of DeepSeek and has had an impact on several sectors of the artificial intelligence business, including AI agents such as AIXBT, the Ai16z framework, the Virtuals AI agent-building platform, and Griffain—a cryptocurrency token that powers a blockchain-based platform.

The cryptocurrency market has also been affected, with artificial intelligence-themed meme currencies such as Fartcoin, as well as tokens like Near Protocol, Internet Computer, Render, and Artificial Superintelligence Alliance, experiencing significant drops. The stock market has been extremely volatile, resulting in a drop of more than $1 trillion in market capitalization for US equities.

Semiconductor makers, which are critical to the infrastructure of artificial intelligence, have suffered substantial setbacks, with Nvidia's stock falling by 16.9%, resulting in a $589 billion loss in market capitalization.

Related News

Concentration camps and the road to exile: Gaza as a laboratory of privatized genocide

Understanding Yemen 2/2: Ansar Allah values

Other large semiconductor makers have also experienced significant drops, with Broadcom and Taiwan Semiconductor Manufacturing Company's stock prices falling 17%. The international response highlights the current ambiguity and unease surrounding DeepSeek's rise, as investors voice concerns about the possible consequences for established firms, as well as the future trajectory of artificial intelligence investments.

How does DeepSeek differ from its peers?

DeepSeek asserts that the production costs of its models are lower than those of the majority of models developed in Silicon Valley. According to estimates from DeepSeek, the manufacturing cost is approximately $5.58 million, whereas the reported expenditure for the development of ChatGPT-4 exceeds $100 million.

“DeepSeek, a less-known Chinese AI company, attracted a significant market reaction because of its affordability by offering its services at only 14 cents per million input tokens—far less than OpenAI's GPT-4 that costs $15 per million input tokens,” Ali Sabir, an AI researcher at Lahore University of Management Sciences (LUMS), told Al Mayadeen English.

He said that this scenario has caused investors to review the financial strategies of well-known artificial intelligence businesses, generating questions about their capacity to remain competitive in a market that values efficiency greatly.

“DeepSeek thrives because it seamlessly integrates innovative training methods and open-source technology into its operations. More people may use it and help it grow easily,” he maintained, adding that through cooperation and knowledge-sharing, the new AI system can reshape the artificial intelligence ecosystem. DeepSeek, he revealed, challenges the notion that achieving dominance in artificial intelligence necessitates substantial infrastructure investment. Its key tokens are "Aha moments" and pure reinforcement learning.

Reactionary measure

Speaking to reporters on January 28, White House Press Secretary Karoline Leavitt mentioned that the National Security Council (NSC) would take a closer look at the potential national security implications surrounding DeepSeek's launch, emphasizing that the administration aims to "ensure American AI dominance."

Since its inception, the United States, Australia, Taiwan, India, Italy, and numerous other nations have prohibited the utilization of DeepSeek on official devices. The United States media asserts that DeepSeek not only undermines user privacy and data security but is also furnished with an integrated real-time content review system designed to disseminate China's official narrative. They contend that this renders it a possible instrument for regulating speech and manipulating public opinion, thereby eliciting considerable vigilance from the international community.

On February 4, the Australian government formally mandated a prohibition on the use of DeepSeek on government apparatus, citing that it presented an "unacceptable security risk." Historically, the Pentagon, NASA, Taiwan's Executive Yuan, and Italy's Garante have enacted comparable prohibitions.

On January 29, US Senator Josh Hawley introduced a legislative proposal aimed at prohibiting American individuals from aiding China in the advancement of artificial intelligence, as well as from downloading or utilizing DeepSeek. The proposed legislation stipulates that offenders would be classified as criminals and could be subjected to a maximum imprisonment of 20 years, in addition to a monetary penalty of $1 million.

In response, the "Jun Zhengping Studio," associated with the Chinese Military Network, disseminated an article on Monday that criticized the United States for employing the "national security" pretext. The article contended that Washington has exhibited an excessive reaction to matters merely associated with the term "China", asserting that it has reached a state of "going astray".

The Voice of America (VOA), a US official outlet that is currently experiencing budgetary cuts under the Trump administration, asserts that the risks associated with DeepSeek extend beyond mere data security concerns. Numerous assessments have indicated that this system's responses markedly align with the "political doctrine" of the Chinese government. In a report published last Monday, the American media organization Wired highlighted that DeepSeek exhibited varying degrees of content censorship during its testing phase.

It is yet to be seen how far the US and its Western allies go against Chinese innovation in the AI sector and how much damage the new Chinese technology could inflict on the US and European tech companies.