DeepSeek Breakthrough Shakes the AI Industry

China’s AI Breakthrough That No One Expected

Chinese company DeepSeek has recently introduced its AI model, DeepSeek-R1, and it seems to shake the tech world. The efficiency of this AI is going to a level where it can be equated with the OpenAI ChatGPT model, even though the development cost was radically different. Although the investment by OpenAI went into hundreds of millions of dollars, DeepSeek spent just $5.6 million. Moreover, DeepSeek did not make use of the very latest chips by Nvidia-which U.S. sanctions have barred from being sold to China-but instead its simplified versions. An important question had thus been asked: if this powerful AI from China could do it at much cheaper costs, are the astronomical expenses by American companies justified?

DeepSeek Crashes the Market

The market reaction was immediate. Nvidia, the world’s biggest supplier of AI chips, saw its stock dive 17% in one day, destroying $600 billion in market value. The Nasdaq 100, comprising the largest technology companies, lost almost $1 trillion in one of the most significant crashes in recent years.

Venture capitalist, engineer, and inventor Marc Andreessen compared this event to the launch of the Soviet Union’s first satellite in 1957. Back then, the world realized that the USSR had overtaken the U.S. in the space race. Now, China has made a similar leap in AI, forcing the West to acknowledge that technological superiority is no longer its exclusive advantage.

The Secret Behind DeepSeek’s Success

Three events are the secret of DeepSeek’s success: The company has managed to sharply reduce the cost of training, while OpenAI spends hundreds of millions on developing AI, DeepSeek managed to achieve the same level of performance for just $5.6 million. This became possible because the company used more affordable Nvidia H800 chips instead of high-end H100 models that competitors relied on. Besides, optimized algorithms allowed DeepSeek to maintain high efficiency at lower computational power.

Another important advantage was DeepSeek’s decision to open-source the artificial intelligence of DeepSeek by releasing its code into the public domain. That means anyone can adapt, change, or enhance the technology. This can be particularly exciting for businesses, researchers, and startups.

A third factor contributing to DeepSeek’s rapid rise to fame was instant and immense popularity among its end-users. Within days, the application of DeepSeek became the most downloaded app in the Apple App Store both in the U.S. and China, beating ChatGPT.

Allegations and Investigations

OpenAI and Microsoft have launched an investigation into the matter, suspecting that DeepSeek used their technologies to train its model. According to them, DeepSeek allegedly used knowledge distillation-a method in which one AI model imitates another to adopt its style of responses and logic. If proved true, DeepSeek will be in serious trouble on legal grounds.

Then there are more US government sanctions for China. At the moment, Beijing cannot purchase the most advanced chips from the US company Nvidia, and tougher restrictions may obstruct further Chinese efforts to develop a new generation of AI models. Yet DeepSeek has managed such a feat thus far, so great is the degree of autonomy in China’s tech industry today.

But despite the accusation, OpenAI CEO Sam Altman said that DeepSeek was a strong competitor. The Chinese model impressed him much as a relative low cost.

“Deep Search” for the AI Industry

DeepSeek is probably going to revolutionize all concepts of what AI should be and how much it should cost. Until recently, there was the belief that the creation of powerful language models requires enormous investments and a highly complicated infrastructure with thousands of state-of-the-art chips. Still, the Chinese company has shown that comparable results are achievable with many fewer resources. This fact is challenging traditional industry practices and forcing major market players to reassess their strategy.

This success might change the balance of power in the technological race. Not long ago, China was considered an outsider in AI, but now it lags behind only a few years. Feeling the threat of falling behind, Donald Trump has already announced the start of the Stargate initiative-the $500 billion program aimed at strengthening the position of American AI developers.

More recently, OpenAI also started to cut prices of its services, part of the company’s desperate attempt to retain users. If DeepSeek were to continue the spree of coming up with open-source solutions for giving fitting competition at almost nil costs, that would raise more ruckus.

On the other side of the coin

But Chinese AI is far from flawless. Answers are filtered to prevent state censorship-the AI declines to respond to inquiries about Xi Jinping or Taiwan. Some analysts question the reported development costs; it likely, they said, the government subsidized electricity, salaries, and computing resources for the project in secret. But it’s difficult to verify that.

What’s Next?

DeepSeek has already broken the game’s rules. From now on, the world’s leading IT companies will have to play by other rules. The technological race will continue, and maybe we really stand at the threshold of a new era in accessible artificial intelligence—or perhaps only at the start of an intense competition between the United States and China in the field of high technologies.

Eesti Firma covers corporate and legal support for all AI projects and assists businesses with being in step with the constantly changing technological and regulatory environments. Feel free to contact us for professional consultations and support!

Thank you!

Your enquiry was successfully recieved!
Our consultants will contact you in the nearest time.