Since its January 10 release, the artificial intelligence application has become increasingly recognized among US users, thanks to its DeepSeek-V3 model, which its developers say “tops the leaderboard among open-source models and competitors the most powerful closed-source models globally,” according to app data research company Sensor Tower.
The achievement indicates how Deep Seek has profoundly affected Silicon Valley, challenging conventional wisdom on US dominance in AI and the efficacy of Washington's export restrictions sought at China's innovative chip and AI talents.
China’s AI Chips Debate
Advanced chips are required to power the training of AI models, ranging from ChatGPT to Deep Seek. Since 2021, the Biden administration has raised the prohibitions designed to avoid these chips from being transferred to China and utilized to train AI models for Chinese companies.
However, in a paper published last month, DeepSeek researchers announced that the DeepSeek-V3 was trained utilizing Nvidia's H800 chips at a cost of less than $6 million.
The assertion that the chips employed were not as powerful as the most innovative Nvidia products Washington has tried to keep out of China, along with the comparatively low cost of training, has caused US tech executives to ask about the efficacy of tech export regulations, even though this detail has since been denied.
The business behind DeepSeek, a little startup based in Hangzhou that was discovered in 2023 when search engine behemoth Baidu unveiled the first Chinese AI large-language model, is not well recognized.
Although numerous Chinese IT companies, both big and small, have since built their own AI models, DeepSeek is the first to get praise from the US tech industry for matching or even outstanding state-of-the-art US models.
Eyes on DeepSeek
All three of Wall Street's major indexes faced loss on Friday, with the S&P 500 falling from a record high because of profit-taking and the negative impact on tech companies after the DeepSeek AI program was launched last week. According to the company, just $5.6 million was utilized to create the model.
The introduction of the program has increased concerns about competition because tech giants like Nvidia, Meta, and Alphabet have spent hundreds of billions of dollars on AI products, which has raised their valuations.
Additionally, it followed Trump's rejection of a new $500 billion project to create artificial intelligence infrastructure in the US. With Tokyo Electron down almost five percent and Advantest down over eight percent, tech and semiconductor companies were one of the major losers in Tokyo as the Nikkei ended in negative territory.
A major backer of Trump's AI initiative, SoftBank, faced a more than eight percent decline. Scale AI CEO Alexandr Wang told CNBC, "We've introduced that DeepSeek is the best performing, or about on par with the best American models."
While Hong Kong saw growth, Shanghai, Bangkok, Singapore, Wellington, Mumbai, and Manila also faced losses. After retreating from record highs on Friday, London and Frankfurt got off to a slow begin, and Paris also faced a decline.