Tech giants like Alibaba and ByteDance and a handful of startups with deep-pocketed investors dominate China’s AI space, making it difficult for small and medium-sized businesses to compete. It’s rare to find a company like DeepSeek that doesn’t have plans to raise capital.
Zihan Wang, a former DeepSeek employee, told MIT Technology Review that while working at DeepSeek, he had access to a wealth of computing resources and was given the freedom to experiment. This was “a luxury rarely available to new graduates at any company.”
In an interview with Chinese media outlet 36Kr in July 2024, Liang said that in addition to chip sanctions, an additional challenge facing Chinese companies is that AI engineering technology tends to be less efficient. “We (most Chinese companies) need to consume twice as much computing power to achieve the same results. Combined with the data efficiency gap, we need up to four times as much computing power.” “It’s possible. Our goal is to continue to close those gaps,” he said.
However, DeepSeek has discovered a way to reduce memory usage and speed up computations without significantly sacrificing accuracy. “The team loves turning hardware challenges into opportunities for innovation,” says Wang.
Liang himself remains highly involved in DeepSeek’s research process, conducting experiments with the team. “The entire team shares a collaborative culture and dedication to serious research,” Wang says.
In addition to prioritizing efficiency, Chinese companies are increasingly embracing open source principles. Alibaba Cloud has released more than 100 new open-source AI models that support 29 languages and address a variety of applications including coding and mathematics. Similarly, startups such as Minimax and 01.AI are also open sourcing their models.
According to a white paper released last year by the China Academy of Information and Communication Technology, a state-affiliated research institute, the number of AI large-scale language models around the world has reached 1,328, of which 36% originate in China. This makes China the second largest contributor to AI after the United States.
“This generation of young Chinese researchers is very aware of open source culture because they have benefited a lot from it,” says Thomas Kitong Cao, assistant professor of technology policy at Tufts University. .
