Foreign media: Lin Jinyang, former head of Tongyi Qianwen technology at Alibaba, said in January this year that the probability of Chinese AI companies surpassing leading American firms within the next three to five years is less than 20%, pointing out that the AI computing power gap between China and the U.S. amounts to "one or two orders of magnitude"—the U.S. controls 74% of global AI computing power, while China holds only 14%.
Optimists argue that Chinese models have already approached U.S. counterparts in multiple benchmark tests, with open-source models occupying nine out of the top ten spots; pessimists, however, worry that Huawei's chip production capacity is merely 2-5% of NVIDIA's, making the computing power disadvantage ultimately insurmountable.
Yet both sides misinterpret the essence of competition: abundant and scarce computing power have given rise to two entirely different innovation models. The U.S., endowed with ample resources, can simultaneously pursue exploration and deployment, whereas China, constrained by limitations, is forced to prioritize efficiency. The crucial question is: does innovation truly emerge from abundance or scarcity?
Original source: toutiao.com/article/1861011998399497/
Disclaimer: The views expressed in this article are solely those of the author.