The Chinese AI app DeepSeek, has quickly risen to prominence, recently surpassing ChatGPT to become the top-rated free app on Apple’s App Store.
This surge in popularity has triggered a selloff in shares of major tech companies, including Nvidia, Microsoft, and Meta, and has negatively impacted European stock markets.
DeepSeek is powered by the open-source DeepSeek-V3 model, which its developers claim was created for less than $6 million—far less than the billions spent by its competitors.
This assertion has sparked debate within the AI community. The app’s emergence challenges the widely held belief that the U.S. is the undisputed leader in AI technology.
As the U.S. imposes restrictions on advanced chip technology exports to China, Chinese AI developers have collaborated and experimented with new methods to continue their work without relying on imported chips.
This has led to the development of AI models that require significantly less computing power and are more cost-effective than previously thought.
Since launching DeepSeek-R1 earlier this month, the company has claimed that its performance is comparable to one of OpenAI’s latest models in areas such as mathematics, coding, and natural language reasoning.
Some industry experts have likened this moment to a “Sputnik moment,” referencing the surprise felt in the U.S. when the Soviet Union launched its satellite in 1957.
The unexpected rise of DeepSeek has startled financial markets, causing significant drops in share prices for companies like ASML and Siemens Energy. Analysts have noted that a low-cost alternative like DeepSeek could threaten the profitability of established rivals, especially given their heavy investments in expensive AI infrastructure.
Despite its success, some analysts caution that challenges faced by Chinese firms could hinder their growth. They suggest that access to advanced chips remains a crucial advantage for U.S. companies.
In response to these developments, a consortium of U.S. tech firms and foreign investors recently announced a major investment initiative called The Stargate Project, which aims to inject $500 billion into AI infrastructure in Texas.
DeepSeek was founded in 2023 by Liang Wenfeng in Hangzhou, China. An information and electronic engineering graduate, Liang also established a hedge fund that backed DeepSeek. He reportedly accumulated a significant stockpile of Nvidia A100 chips before export restrictions were imposed on China. By combining these chips with more affordable alternatives still available for import, he was able to launch DeepSeek successfully.
In a recent interview, Liang expressed surprise at how sensitive pricing had become in the market, stating that his team focused on calculating costs and setting prices accordingly without anticipating such a strong reaction.
Source: BBC
