Baidu, Alibaba and Tencent initiatives will help China ‘aggressively’ narrow AI gap with US
Government funding and regulatory support, deep pool of AI talent and large amounts of data gathered from businesses will boost efforts
The combined initiatives of Chinese internet powerhouses Baidu, Alibaba Group Holding and Tencent Holdings are helping the mainland chip away at the lead held by the United States in artificial intelligence (AI) innovation globally.
“There’s still a huge gap between the [research and development spending of] Chinese internet companies and their global peers,” John Choi, the head of China internet research at Daiwa Capital Markets, said on Tuesday.
Choi, however, pointed out that ample government support in terms of regulation and research funding, deep pools of AI and engineering talent, private sector investment and large amounts of digital data for deploying machine-learning techniques to existing products would enable the mainland companies “to narrow that gap quite aggressively over the coming years”.
That also indicates how Baidu, Alibaba and Tencent are likely to continue being well entrenched in the mainland economy through AI after dominating the fields of online search, e-commerce, and online games and social media, respectively, in the country.
With more than 60 years of international development behind it, AI is an umbrella term covering several technologies that include machine learning, cognitive computing, natural language processing and so-called neural networks. Machine learning, for instance, is a type of AI focused on computer programs that have the ability to learn when exposed to new data.
In August, the Chinese government announced its national development plan, which included building up a highly competitive AI industry that would be worth 1 trillion yuan (US$150.7 billion) by 2030.
Choi said AI was already deeply rooted in many core internet services on the mainland. “From search engines to content recommendations, Chinese internet companies are folding machine-learning technology into their services with the aim of improving information discovery and monetisation efficiency,” he said.
Nasdaq-listed Baidu, which is considered by Daiwa as one of the pioneers in AI research on the mainland, has sharpened its focus on leading the global market for autonomous driving technology.
In July, Baidu said more than 50 companies, including 13 Chinese car makers and two western automotive giants – Ford and Daimler – as well as component providers and ride-sharing operators such as China’s UCAR and Singapore-based Grab had joined Project Apollo, the company’s open autonomous driving platform.
“We anticipate there being a large addressable market for Baidu’s autonomous driving business, though we believe it is unlikely to bring in significant revenue in the next three years,” Choi said in a recent Daiwa report.
New York-listed Alibaba, which owns the South China Morning Post, announced last month that it would invest US$15 billion over the next three years into “cutting-edge technology”, including AI, through its global research effort called Discovery, Adventure, Momentum and Outlook (Damo) Academy.
This planned investment will more than double the US$6.4 billion Alibaba spent on research over its past three financial years, according to data from Bloomberg.
Alibaba, which operates massive online retail platforms Taobao Marketplace, Tmall.com and AliExpress, said the Damo Academy will establish new laboratories on the mainland, Singapore, Israel, the US and Russia.
“We believe Alibaba has a natural edge in AI research that has yet to be fully appreciated,” said Choi. “We see multiple use cases for machine learning in Alibaba’s existing product range.”
Shenzhen-based Tencent, the world’s largest games company by revenue and China’s biggest social media operator, said in August it would boost investment in AI.
Tencent set up an AI research lab in Seattle in May, led by former Microsoft scientist Yu Dong. It also hired more than 50 AI researchers, backed by about 200 engineers, on the mainland.