US analysis of AI paper ranking overstates China's influence: analysts

By Chen Qingqing Source:Global Times Published: 2019/3/14 22:08:40

US analysis overstates China’s influence in sector

A robot dances at an industry expo in Beijing. File photo: VCG

Don't get too excited about a report suggesting China has overtaken the US in artificial intelligence (AI) research, as China still lags the US in fundamental research in the sector, Chinese analysts said on Thursday. 

The Seattle-based Allen Institute for Artificial Intelligence, a research institute founded by US technology giant Microsoft co-founder Paul Allen, said on Wednesday that China surpassed the US in published AI papers, by analyzing more than 2 million AI papers published through 2018. 

While taking a look at the top 10 percent of the most cited papers - a measure of high-impact AI research - China has risen, with a share of 26.5 percent in 2018, the institute said. Meanwhile, the share of papers published in the US declined to 29 percent, and the trend is continuing. 

The analysis also showed that China is poised to overtake the US in the most-cited 50 percent of papers this year, in the most-cited 10 percent of papers next year, and in the 1 percent of most-cited papers by 2025.

"Government support, an active research and development (R&D) environment in China, as well as a massive return of oversea talent," are reflected in the findings, Xiang Yang, an industry analyst at Beijing-based CCID Consulting, told the Global Times on Thursday. 

Chinese companies have become active players in leading in the AI field, and some of them have teamed up with domestic universities to come up with new efforts in AI-related technologies such as neuro-linguistic programming and machine learning.  

For instance, Chinese voice recognition technology provider iFlytek recently established an affiliate with the South China University of Technology to focus on brain-machine interface research, according to the company's statement in February. This effort will deepen collaboration between the company and the college in core AI-related technologies including brain-machine interaction and hybrid intelligence. 

However, the increasing market share of Chinese AI research papers on a global scale does not necessarily mean a far-reaching influence on the industry, analysts said. 

"As far as we know, we have limited impact in terms of AI-related technologies, as well as research and algorithms. Many companies take foreign-published algorithms as a reference, and put them into practice," Lu Yanxia, a senior research manager for IDC China, told the Global Times. 

US-based Google said in September 2018 that its AI development framework TensorFlow has been downloaded more than 2 million times since 2015, and Chinese companies such as Inc, Xiaomi and NetEase are among the users.

Still, some domestic AI companies have been establishing open-source models to help increase the use ratio of technologies, indicating ascendency of the overall industry, Lu noted. 

Newspaper headline: Nation rises in AI paper ranking

Posted in: ECONOMY

blog comments powered by Disqus