The International Data Corporation recently released a report on the AI digital human market in China, which outlines current digital human applications, and predicts that China’s AI digital human market size will reach RMB 10.24 billion by 2026.
In the report, digital humans have been categorized into 5 levels, with L1 being “having digital appearances but relying on humans to make decisions”, and L5 being “capable of fully intelligent interactions, autonomous decision making, and autonomous task execution”. Currently, digital human products on the market are mostly restricted to L1 and L3 and applied in industries such as customer service, finance, entertainment, tourism, and governments.
The IDC report also showcases several key digital human commercialization cases, including ones at the Shanghai Pudong Development Bank(SPDB). The bank’s digital employees are provided by Baidu AI Cloud since 2019. SPDB currently has large-scale applications of digital employees for customer services, document review, internal training, and sales.
85% of customer service calls are answered by digital humans at the SPDB. They also make 500-600 thousand calls while human employees only make 20 thousand calls, per the IDC report. In addition, document review digital employees can provide productivity equivalent to 200-300 human staff/year, and human customer service staff and financial managers trained by AI digital staff have increased the average amount of sales by 10-20%.
Baidu, SenseTime, and Microsoft’s spinoff Xiao Ice have been listed as “leading first-tier companies” in the industry, with Baidu AI Cloud ranking first in terms of “AI capabilities” and “user experiences”, and Xiao Ice ranking first in terms of “market and ecological capabilities” and “commercialization cases”.
Xiling, Baidu’s AI digital human platform, has 98% accuracy in speech accuracy and 98.5% accuracy in mouth shape synthesis.
Baidu’s AI digital humans benefit from its own AI model Plato, which has ten billion parameters, making it the world's largest Chinese and English dialogue generation model. With Plato, Xiling digital humans are able to have logical, informative, and interesting multi-turn conversations with users in English and Chinese.
In addition, Baidu’s digital humans also cater to underserved communities such as the deaf. According to Baidu, its sign language digital human has helped events to stream to over 27 million hard-of-hearing users, through translating and narrating.
Favorable policies, application value, market demand, capital expenditure, and technology maturity are combined to drive the rise of the AI digital human market, according to the IDC report. Virtual anchors and animated sign language should be promoted and widely used in the production of news broadcasting and other TV programs, according to a "Fourteenth Five-Year Plan" published by China’s National Radio and Television Administration, emphasizing the use of science and technology in radio, television, and network.
“Today's AI digital people are already showing clear commercial value in a number of areas," said Yanxia Lu, an associate research director for IDC China's emerging technology research, adding that “In the future, there definitely will be large-scale teams of digital humans co-existing with humans in life and workplaces. Industry users can start introducing AI digital employees in relatively mature application scenarios, and also remain patient with these application scenarios and not set their expectations too high.”
Cover image provided by Baidu.