Elon Musk predicts the ultimate three strong AI players: xAI, Google, and the Chinese national team

Elon Musk predicted in his "Moonshots" podcast that in the future, three major forces will dominate the AGI race: his xAI, Google, and the "Chinese national team". This ultimate showdown focused on computing power has reached a critical turning point.

During the two-hour conversation, Musk clearly outlined the final outcome of AGI competition: AI competition is a knockout race where the strong always remain strong, and only a few entities with outstanding core computing power and execution ability can reach the top. Among the three major forces he named, the inclusion of the "Chinese national team" stems from his recognition of China's AI execution capability - he clearly stated that China's AGI computing power reserves will "far exceed" other regions in the coming years and become a global core force.

Musk's prediction goes straight to the core: the essence of the computing power battle is a competition between electricity and chip production capacity, which are the "underlying cornerstones" of AI development. He predicts that by 2026, China's electricity output will triple that of the United States, and this energy advantage will provide solid support for China's AGI training, giving the "Chinese national team" the confidence to compete on the same stage as xAI and Google.

As Musk's core layout, xAI highlights its competitiveness. SpaceX has fully acquired xAI and plans to deploy orbital data centers using space solar energy, making space a low-cost computing location; At the same time, xAI has completed a $20 billion financing to build the world's largest GPU cluster, and the iteration speed of the Grok model is gradually narrowing the gap with Google.

As a veteran AI powerhouse, Google's position is difficult to shake. With profound technological accumulation, sufficient chip production capacity, and a mature algorithm ecosystem, its DeepMind laboratory continues to lead the industry's development. Musk also acknowledges that Google is the strongest competitor of xAI, possessing core AI technology and ecological resources, which is the goal that xAI needs to strive to surpass.

Compared to xAI's "latecomer" and Google's "deep foundation", the advantages of the "Chinese national team" are mainly reflected in the rapid improvement of computing infrastructure and the dual empowerment of policies and markets. In recent years, China has continued to increase its investment in the field of AI computing power. The national supercomputing Internet core node has been put into trial operation. Supported by Dawning Wanka super cluster system, it can provide more than 30000 domestic AI computing power to the outside world, build the country's largest domestic AI computing power resource pool, and provide efficient services for large-scale AI computing such as trillion parameter model training, high-throughput reasoning. By the end of 2025, the national supercomputing Internet has served more than 1 million users, and has supported 196 million operations, significantly reducing the cost of computing and the threshold for use.

At the same time, China's independent breakthroughs in the chip field are also accelerating. Data shows that China's self-sufficiency rate of artificial intelligence GPUs has increased from less than 10% in 2020 to about 34% in 2024, and is expected to rise to about 82% in 2027. Although there is still a gap between domestic GPUs and international flagship products in terms of absolute computing performance and energy efficiency, with the continuous improvement of the domestic chip supply chain and the continuous improvement of computing resource scheduling efficiency, China is gradually filling the chip self-sufficiency gap and laying a solid foundation for the continuous expansion of computing power reserves. The power advantage predicted by Musk is more in line with China's rapid development in the clean energy field in recent years - China's leading position in solar deployment, battery energy storage and other fields is providing new possibilities for the green and sustainable development of AI computing power.