The Chinese authorities, according to online sources, have developed an AI Industry Development Action Plan program aimed at the comprehensive development of AI infrastructure and services in the country. It is planned to allocate at least ¥1 trillion, or about $138 billion, for these purposes over the next five years.
It is noted that the AI Industry Development Action Plan initiative is the PRC’s response to the large-scale Stargate project for the development of AI in the United States. In the case of Stargate, we are talking about creating a joint venture between OpenAI, Softbank and Oracle, which will develop “physical and virtual infrastructure to support the next generation of AI.” The project will involve Microsoft, Arm and Nvidia, as well as the MGX fund from the Middle East. The total investment is expected to reach $500 billion.
The AI Industry Development Action Plan is reportedly supported by Bank of China. The plan includes four main directions. The first is the strengthening of national technological independence: this is especially important in conditions of intensifying sanctions by the United States. The Bank of China will provide comprehensive support to enterprises engaged in innovative technologies, including AI.
Image Source: unsplash.com/Dominic Kurniawan Suryaputra
The second direction is the development of AI infrastructure, including increasing computing power, building new data centers and auxiliary facilities. The third strategic goal is to accelerate innovation in the field of AI, which involves the development of specialized applications. The fourth direction is promoting the use of AI in various scenarios, including in the field of robotics, advanced materials, biomanufacturing, etc.
The ultimate goal of the AI Industry Development Action Plan is to create an AI-centric financial ecosystem to accelerate innovation, enable industrial development and sustainable development.
It is worth noting that the Chinese AI startup DeepSeek recently caused a stir in Silicon Valley: the company released the “reasoning” AI model R1 and made it completely free for everyone to use. Moreover, DeepSeek has published instructions on how to cost-effectively build a model that can independently learn and improve without human supervision. This has challenged many Western competitors, including OpenAI.