The first working day after the Lunar New Year break at DeepSeek began where the Chinese start-up left off last week, with more plaudits heaped on its highly touted feat of building powerful and more cost-efficient artificial intelligence (AI) models in spite of US tech restrictions.
Advertisement
But DeepSeek’s head office in Hangzhou, capital of eastern Zhejiang province, on Wednesday kept mum on the firm’s potential next moves, as it declined all requests from news media, including the South China Morning Post. More security personnel were deployed outside its office building on the same day, according to one of the guards on site who declined to be identified.
The property management at its building ushered all uninvited visitors to a room to turn down their requests for a visit. The entrance to DeepSeek’s offices on the 12th floor of the building was locked, while a peek through its glass doors showed holiday decorations scattered on the floor.
According to a source familiar with the company, DeepSeek was as surprised as others on the impact of its two new open-source models, DeepSeek-V3 and DeepSeek-R1, in the global AI market.
On Wednesday, the Communist Party secretary of southern Guangdong province, Huang Kunming, commended DeepSeek for taking on the United States’ AI titans with “courage and vigour”. Huang, a member of the Politburo, is the first senior Chinese government official to publicly praise the company.
Advertisement
In a social-media post, renowned Chinese life scientist Rao Yi said DeepSeek’s AI breakthrough marked the country’s greatest contribution in terms of technology since the first opium war in the mid-19th century, without elaborating.
