News

Deepseek AI breakthrough sparks market disruption, challenges chip supply chain dominance as China’s innovative ecosystem fuels advancements, raising questions about the future of AI development, shifting focus toward efficiency and open source.

Over the last two years, the tech industry has enjoyed an insatiable appetite for Nvidia’s chips, but it has now been dealt a harsh reality check.

A little-known Chinese artificial intelligence startup called DeepSeek has demonstrated it could do more with less of what Nvidia produces. The development has caused tech stocks to tumble, forcing giant companies like Meta and Nvidia to answer tough questions about their future. Still, it has also raised questions about the nature of AI development, its costs, and the strategic importance of resources like advanced computer chips.

DeepSeek’s emergence has challenged the belief that building powerful AI systems requires massive investments in specialized AI chips.

This has disrupted the notion that only the biggest tech companies could afford to develop leading AI technologies. “It is difficult to know exactly how to make money on AI,” said Mike Ogborne, founder of Ogborne Capital Management, a hedge-fund firm in San Francisco that oversees a position in Nvidia. “This could be the first day of a lot more pain,” he added, hinting at the potential for further market corrections.

New paradigm for AI development

DeepSeek’s accomplishment lies in its ability to create a powerful AI model with far less computing power than previously thought necessary. While top companies typically train their chatbots with supercomputers using as many as 16,000 chips or more, DeepSeek’s engineers stated they needed only about 2,000 Nvidia chips. 

This breakthrough resulted from using fewer chips and the development of more efficient data analysis methods. DeepSeek’s research paper laid out clever and impressive methods for building AI technology with fewer chips, which other companies can adopt, thereby accelerating the overall improvement of AI models.

It has led to reconsidering the relationship between computing power and AI capabilities. Before DeepSeek’s breakthrough, there was a consensus in the tech industry that AI advancements required more computing power.

The prevailing belief was that “bigger chips equal bigger A.I. capabilities,” leading to a heavy reliance on specialized AI chips, primarily from Nvidia, which controlled a large portion of the market and required massive investments in data centers.

DeepSeek has proven they could build powerful AI without enormous hardware outlays, prompting investors to question the value of AI infrastructure and hardware.

Impact on the semiconductor supply chain

DeepSeek’s success has significant implications for the supply chain, especially for companies like Nvidia, which had soared to the top of the stock market by selling the computer chips fueling the world’s artificial intelligence boom. “DeepSeek has levelled the playing field,” said Stephen Yiu, chief investment officer of Blue Whale Growth.

With the understanding that powerful AI systems can be developed with fewer chips, demand may shift away from the most advanced and expensive chips. It could potentially lower costs and make AI development more accessible to a wider range of players.

The change also challenges the notion of U.S. dominance in AI development and the effectiveness of trade restrictions.  To maintain U.S. leadership in the global AI race the Biden administration implemented rules limiting the powerful chips sold to China and other rivals.

DeepSeek, however, achieved its breakthrough despite these restrictions, demonstrating that Chinese researchers are finding ways to innovate with the available resources.

The achievement has raised questions about the unintended consequences of trade restrictions, as such controls may be forcing researchers in China to get creative with freely available tools on the market.

The open-source nature of DeepSeek’s AI system also stands to disrupt the established order of the AI supply chain. DeepSeek enables others to build and distribute their products using the same technologies by sharing the underlying computer code with other businesses and researchers. This open-source approach has helped DeepSeek and other Chinese companies build competitive AI systems quickly and inexpensively.

China’s ecosystem and AI development

DeepSeek’s rapid rise is also a testament to China’s unique approach to AI development. This approach relies not solely on advanced chips but on a combination of cultural factors, abundant data, and an entrepreneurial spirit.

According to Kai-Fu Lee, a prominent figure in AI and the author of AI Superpowers, China’s AI success is due to several factors. China’s culture encourages imitation and adaptation but also promotes a strong work ethic and entrepreneurial drive. 

“The pace is incredible in China,” Andrew Ng, the deep-learning pioneer who founded the Google Brain project and led AI efforts at Baidu, told Lee. “While I was leading teams in China, I’d just call a meeting on a Saturday or Sunday, or whenever I felt like it, and everyone showed up and there’d be no complaining,” he added.

In his book, Lee also points out that the vast amounts of data generated by China’s unique internet ecosystem and a proactive government policy environment make China a significant force in the AI landscape. The Chinese government has actively promoted mass entrepreneurship and innovation through policies and funding. Ambitious mayors across China are scrambling to turn their cities into showcases for new AI applications. 

“When Chinese investors, entrepreneurs, and government officials all focus in on one industry, they can truly shake the world,” said Lee. “Indeed, China is ramping up AI investment, research, and entrepreneurship on a historic scale.”

While the U.S. political system may aggressively punish missteps or waste funding technological upgrades, China’s techno-utilitarian approach rewards proactive investment and adoption.

“It’s a constant pace of decision-making,” said Ng. “The market does something, so you better react. That, I think, has made the China ecosystem incredible at figuring out innovations and how to take things to market,” he concluded.

China’s approach shines through a focus on implementation rather than just basic research and a willingness to embrace real-world applications of AI.

Chinese companies are known for their “going heavy” approach. They immerse themselves in the messy details of food delivery, car repairs, and shared bikes, allowing them to gather data on their users’ real-world habits. This data advantage enables Chinese companies to compete with Silicon Valley juggernauts that tend to focus on online interactions.

A changed future for AI

DeepSeek’s emergence has raised many questions about the future of AI and its supply chain. The company’s success has shaken the consensus that building bigger and better AI systems requires enormous financial investments in specialized AI chips and data centers. As Marc Andreessen, a tech venture capitalist, posted on social media, DeepSeek is “AI’s Sputnik moment,” indicating a potential shift in the global AI landscape. 

“There hasn’t been a winner here,” said Daniel Newman, chief executive of Futurum Group, reflecting the uncertainty in the market. “The tools have been just OK. But if everyone can make better models at lower costs, then people may start adopting AI,” he concluded.

While some chip-industry insiders don’t believe DeepSeek will supplant AI’s incumbents or that its claims of needing small amounts of computing power to create powerful AI models mean Nvidia’s business is doomed, the company has undoubtedly disrupted the market and challenged established norms. 

The future of AI may involve more efficient resource use, innovative technical approaches, and a willingness to challenge existing assumptions rather than simply scaling up existing systems with massive computing power.

DeepSeek’s ability to achieve significant results using fewer resources has set a precedent that could reshape the future of AI development and the chip supply chain that drives it.