A Chinese startup, DeepSeek, has made a significant splash in the AI world. Its AI Assistant app has recently overtaken ChatGPT to become the top-rated free app on the Apple App Store in the United States.
DeepSeek’s success is attributed to its powerful DeepSeek-V3 model, which the company boasts rivals the performance of the most advanced AI models globally, including closed-source ones. Since its launch on January 10th, the app has experienced rapid growth in popularity among US users, according to app research firm Sensor Tower.
This achievement has sent shockwaves through Silicon Valley, challenging the long-held assumption of US dominance in the AI field. It also raises serious questions about the effectiveness of the US government’s export controls designed to limit China’s access to advanced chips and AI technology.
AI models like ChatGPT and DeepSeek require powerful chips for training. The Biden administration has tightened restrictions on the export of these high-end chips to China, aiming to hinder their AI development. However, DeepSeek researchers recently claimed that they trained the DeepSeek-V3 model using Nvidia’s H800 chips, a less advanced but still powerful chip, at a surprisingly low cost of under $6 million.
While this claim has been met with some skepticism, it has ignited debate within the US tech industry. The possibility that DeepSeek achieved such impressive results with less advanced chips and a relatively low training budget casts doubt on the effectiveness of the current export control measures.
DeepSeek, a young company founded in 2023, has quickly emerged as a major player in the Chinese AI landscape. While numerous Chinese tech giants have released their own AI models, DeepSeek stands out as the first to garner significant praise from the US tech industry for its ability to match or even surpass the performance of leading American models.
This development underscores the rapidly evolving global AI landscape and the increasing competitiveness of Chinese AI technology.
