Recently, the Xinhua News Agency in China unveiled at the World Internet Conference a new sharply dressed news anchor, that looked like Qiu Hao, a well known news presenter. However, the news agency described the new personality as the world’s first artificial intelligence (AI) news anchor that was modeled on the likeness of Qiu Hao. The AI anchor was produced with facial landmark localization and face reconstruction.
The video of this AI powered personality who started by saying “This is my very first day at Xinhua News Agency . . . I look forward to bringing you the brand new news experiences.” The anchor designed jointly with the Chinese search engine company Sogou.com “learns from live broadcasting videos by himself and can read texts as naturally as a professional news anchor.”
Xinhua points to what it sees as certain advantages of an AI anchor, saying it “can work 24 hours a day on its official website and various social media platforms, reducing news production costs and improving efficiency.” South China Morning Post suggests it could save networks money in news anchor salaries, and even “one day challenge the human variety.”
As a society we have seen automation of many traditional jobs that required physical labor. The industrial age scaled mechanization as a way to extend human performance by expediting the production of objects or processes that enhanced human activity, or that humans were needed to enhance the machine. This caused increases in efficiency and industrial output, which also increased profitability and wealth. However, many trades were phased out due to automation and the rise of unions was a response to unhealthy and exploitative working conditions as well as by issues around automation.
In today’s day and age we are seeing major transformations through digital automation which is starting to affect while collar professions like accounting, law and medicine. Intellectual capital is being automated and infinitely reorganized to cut through the clutter and tailor content to meet specific goals. This will allow greater efficiencies, but will introduce new challenges of what is the role of humans and the definition of work.
Xinhua is capitalizing on what is called “automated journalism” that interprets, organizes, and presents data customized to fit a certain voice, tone, or style. These AI systems are still in their infancy, but through continual iterations and refinements in shorter periods of time they will become more sophisticated, feel more “natural” and become accepted by humans. Since this AI is a human metaphor, it will undoubtedly cause new types of empathy from humans that interact with these rich avatars.
We have also seen this technology used to create modified videos where another person’s lips and voice is seamlessly grafted into existing video, changing the meaning of the original intent of a video. While this process is intentionally modifying a video to change the meaning of it, and the Xinhua News Agency’s intent is to create an automated newscaster service, the results could be the same – a distorted message with a veneer of authenticity.
As game-changing technological innovations like these become more common place, the question arises of what is authenticity and human judgement. With increasing acceptance of automated content and engineered experiences that anthropomorphize media, our ability to distinguish “authentic” from “modified” content will become difficult and could lead to intentionally manufactured conflicts. This would meet Xinhua’s statement of “one day challenge the human variety” of direct experiences that are captured by media to be shared – to modified and manufactured experiences that were never directly captured by media in the analog world.