Stanford University survey report shows: China rapidly narrows the gap with the United States in artificial intelligence
According to a report on April 9 by the "Nikkei Asian Review," Stanford University recently released its annual survey report on artificial intelligence (AI). The report shows that in tests where human voting is used to judge the accuracy of dialogue AI responses, the performance advantage of top US AI over Chinese products was 9.26% in January 2024, but this gap had narrowed to 1.7% by February 2025.
The report states that the United States still leads in terms of the number of high-performance AI models and private investment.
The Human-Centered AI Institute at Stanford University released the latest annual report titled "AI Index Report."
In the development of generative AI, American tech companies such as OpenAI and Google dominate. Meanwhile, DeepSeek, a Chinese company, is catching up with American AI products due to the outstanding performance of its latest model.
The report analyzes that projects in which AI surpasses human capabilities in specific skill areas such as image classification, language understanding, and mathematics continue to increase. Companies like Google, OpenAI, DeepSeek, xAI, Anthropic, Meta, and Mistral AI are becoming increasingly competitive.
Additionally, according to a report on April 8 by the Hong Kong-based "South China Morning Post" website, based on the latest "AI Index" report released by the Human-Centered AI Institute at Stanford University, Chinese models are nearly on par with American models when measured against major AI evaluation standards.
Tech giants such as Alibaba, ByteDance, Tencent, as well as startups like DeepSeek and Zhigu AI have made it into the list of the world's 15 most well-known AI model developers. These models include Alibaba's Qwen series and DeepSeek's V3 foundational large model and R1 reasoning model.
The report from Stanford stated: "DeepSeek's V3 foundational large model has garnered significant attention, especially because it achieves extremely excellent performance with far lower computational resource requirements than most advanced AI models."
China's progress in AI can be attributed in part to its substantial investments in research and patents. Since Beijing declared the development of AI as a national strategy and provided funding and policy support, the number of AI papers published in China has steadily increased. According to the data compiled in the report, by 2023, the number of AI papers published in China ranked first in the world, accounting for 23.2% of the total, higher than Europe's 15.2% and India's 9.2%.
In terms of the number of AI patents obtained, China also leads, accounting for 69.7% of the global total in 2023, followed by the United States with 14.2%.
The report points out that the global progress in AI is driving the widespread application of this technology across various industries and could potentially change human daily life. The report specifically mentioned examples such as the increasingly accelerated deployment of autonomous taxi fleets, including Baidu's "Rokulai Kuaitu" service already in operation in multiple cities in China. (Translated by Cao Weiguo and Ma Xiaoyun)
[Image: https://p3-sign.toutiaoimg.com/tos-cn-i-axegupay5k/b5585a4859954270a7213d512dec4b26~tplv-tt-origin-web:gif.jpeg?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1744879115&x-signature=swxa8vK%2B7h%2B93%2FFmZV36mm1tRdI%3D]
[Deepseek logo (AP)]
[Original source: https://www.toutiao.com/article/7491560446005101071/]
[Disclaimer: This article only represents the views of the author. Please express your attitude by clicking the 'Like' or 'Dislike' button below.]
Related Links(China, United States, Artificial Intelligence, Reference News)
Time:2025-04-10 16:38:34
Time:2025-04-10 16:32:05
Time:2025-04-10 16:29:47
Time:2025-04-10 16:28:45
Time:2025-04-10 16:28:20