US President Donald Trump has described the ascent of Chinese firm DeepSeek as “a wake-up call” for the American technology sector, following the debut of its artificial intelligence (AI) model that sent ripples through Wall Street.
Major tech firms, including Nvidia, experienced a significant decline in their share prices. The chip giant lost nearly $600 billion (£482 billion) in market value.
DeepSeek’s assertion that its R1 model was developed at a significantly lower cost than its competitors has sent ripples through the industry, prompting a reevaluation of the future of AI leadership in the United States and the extent of investments that American companies are considering.
In a remarkable achievement, DeepSeek has surged to the top of the charts as the most downloaded free app in the United States, accomplishing this feat just one week after launch.
In reaction to recent updates, Trump remarked that advancements in China’s AI sector could potentially benefit the United States.
“If it were possible to achieve the same outcome at a lower cost, would you consider it?” “I believe that’s a positive development for us,” he stated to reporters while aboard Air Force One.
He expressed confidence regarding the breakthrough, asserting that the United States will remain a leading force in the sector.
DeepSeek operates on the open-source DeepSeek-V3 model, which its researchers assert was developed at a cost of approximately $6 million (£4.2 million). This figure stands in stark contrast to the billions invested by competing entities. This assertion has faced challenges from various stakeholders within the AI community.
This development’s emergence coincides with the United States implementing restrictions on the sale of advanced chip technology, which is essential for artificial intelligence, to China.
In the face of inconsistent access to imported advanced chips, Chinese AI developers have resorted to collaboration and experimentation, sharing their findings and exploring innovative methodologies in the field.
The development of AI models has significantly reduced the computing power needed compared to previous iterations.
The implications suggest that these costs are significantly lower than earlier estimates, a development that could disrupt the industry landscape.
In the wake of a market shock in the United States on Monday, the FTSE 100, which tracks the performance of the UK’s largest publicly-listed companies, demonstrated notable resilience in early trading on Tuesday, climbing by 0.46%.
Futures for the tech-focused Nasdaq index increased modestly by 0.1%, following a slight uptick in Nvidia stock during after-hours trading.
Shares in Japanese AI-related companies, such as Advantest, Softbank, and Tokyo Electron, experienced a significant decline, contributing to a 1.4% drop in the benchmark Nikkei 225 index.
Numerous markets across Asia have shut down in observance of the Lunar New Year holiday. Mainland China’s financial markets are set to close on Tuesday, with plans to resume operations on 5 February.
The founder of DeepSeek remains a topic of interest among industry observers.
Founded in 2023 by Liang Wenfeng, the company is based in Hangzhou, a city located in southeastern China.
A 40-year-old information and electronic engineering graduate established the hedge fund that provided financial support for DeepSeek.
Recently, he was spotted attending a meeting that brought together industry experts and Chinese Premier Li Qiang.
During a July 2024 interview with The China Academy, Mr. Liang expressed his astonishment at the response to the earlier iteration of his AI model.
“The sensitivity surrounding pricing caught us by surprise,” he remarked.
“The team maintained its rhythm, carefully assessing expenses and adjusting prices in response.”
Following the launch of DeepSeek-R1 earlier this month, the company claimed that its performance is comparable to one of OpenAI’s latest models, particularly in mathematics, coding, and natural language reasoning.
DeepSeek’s technology has garnered acclaim from notable figures, including OpenAI’s chief, Sam Altman. He described it as “an impressive model, particularly around what they’re able to deliver for the price,” while also noting that OpenAI is expected to “obviously deliver much better models” in the future.
“DeepSeek’s capacity to compete with US models, despite restricted access to cutting-edge hardware, illustrates how software innovation and data efficiency can offset hardware limitations,” stated Marina Zhang, an associate professor at the University of Technology Sydney, specializing in China’s high-tech sectors.
Ion Stoica, co-founder and executive chair of the AI software firm Databricks, spoke to the BBC about how DeepSeek’s reduced cost may encourage more companies to integrate AI into their operations.
“Should this occur, the decrease in expenses could expedite advancements in AI,” he stated. The market is poised for rapid expansion, with its value expected to increase at an accelerated pace.
A Chinese company asserts that its model can be trained using 2,000 specialized chips, a significant reduction compared to the estimated 16,000 required for leading models in the industry.
However, skepticism remains among specific individuals. Concerns have been raised regarding confident assertions by DeepSeek, with notable figures such as tech mogul Elon Musk expressing skepticism.
In response to a post asserting that DeepSeek possesses approximately 50,000 Nvidia chips that are now prohibited from exporting to China, he remarked, “Obviously.”
