Tencent, a latecomer to AI field with a slight edge


ANN@The China Post

Tencent Holdings Ltd, best known for its all-in-one app WeChat, is making inroads into artificial intelligence or AI to give its exisiting businesses a tech fillip.

Though a latecomer to the voice computing field, the Shenzhen-based company aims to catch up with existing players on the back of its sprawling ecosystem and increasing investments in speech recognition and natural language processing capabilities.

In late June, Tencent unveiled Xiaowei, an intelligent voice computing platform developed on the basis of WeChat’s voice recognition technology.

Xiaowei comes with all the usual bells and whistles of voice assistants: weather reports, traffic updates, music requests and news blurbs.

Tencent’s Chinese peers Alibaba Group Holding Ltd and Baidu Inc had also launched similar open voice computing platforms, in the hope that more third-party service partners will connect with their digital assistants.

What helps Tencent stand out, however, is its sprawling ecosystem. It has interests in e-commerce, ride-hailing, food delivery, education and search engine businesses.

It has licensing deals across mobile/online games, movies, music, sports and literature. All of these can be linked to Xiaowei and used through voice commands by users of Tencent’s products and services.

WeChat’s 938 million users share voice messages with friends. That’s data fodder for Xiaowei, which can improve the accuracy of voice recognition.

Mao Hua, head of the Xiaowei project at Tencent, said among the current partners are ride-hailing company Didi Chuxing, electronic appliances brand Midea and search engine Sogou.

Hardware maker Asus also plans to launch Xiaowei-powered educational and home companion robots this September.

The Xiaowei platform came into being after Tencent established its first US-based AI research laboratory in Seattle in May.

Former Microsoft researcher Yu Dong is in charge of the lab, which focuses on how to drive fundamental research in, and commercial application of speech recognition and natural language understanding.

Yu said the development of AI depends on four core elements: rich application scenarios, big data, strong computing ability and excellent scientific talents.

“In the past 18 years, Tencent has accumulated rich resources in these aspects, which are important basics to develop AI,” Yu said.