Accessibility links

Breaking News

Future of News: Will Robot Readers Take Over?


Xinhua news anchor Qiu Hao stands next to an AI virtual news anchor based on him, at a Sogou booth during an expo at the fifth World Internet Conference in Wuzhen town of Jiaxing, Zhejiang province, China, Nov. 7, 2018.
Xinhua news anchor Qiu Hao stands next to an AI virtual news anchor based on him, at a Sogou booth during an expo at the fifth World Internet Conference in Wuzhen town of Jiaxing, Zhejiang province, China, Nov. 7, 2018.
Future of News: Will Robot Readers Take Over?
please wait

No media source currently available

0:00 0:06:46 0:00

Technology is forever changing the way we get our news. Many people now get a lot of their news on electronic devices, instead of traditional media, such as newspapers, television or radio. An increasing number of people also report turning to social media for information.

Now, there is a new technology that could greatly change the way we get the news: computer-created news readers.

Recently, China’s state-supported news agency Xinhua launched what it called the world’s first artificial intelligence, or AI news presenters. Xinhua said it worked with the Chinese search engine Sogou to develop the AI presenters.

The news readers - created by machine learning technology - are based on two real-life Chinese newsmen. One is able to present newscasts in English; the other, in Mandarin Chinese.

In Xinhua’s report about the launch, it said machine learning was used to examine video images and sounds of the two newsmen. The system then created the AI robotic presenters, which look and sound like real people.

Similar technology has been used to create video products known as “deepfake videos.” A deepfake is a video that looks real, but was electronically changed. Such videos can make people appear to say things they never said or do things that never actually happened.

Xinhua said the AI presenters will be a great improvement to the news team because they “can work 24 hours a day” on its website and on social media. One of the presenters even promised to “work tirelessly” to keep people informed.

Xinhua claimed the AI newsmen were able to read the news “as naturally as” a human presenter.

Some machine learning experts said the system showed off China’s latest progress in voice recognition, text-to-speech technology and data analysis. But several experts suggested that the term AI does not correctly describe abilities of the robotic news readers.

On social media, many Chinese noted that the AI presenters did not seem real. People criticized them for not being more lifelike. On China’s Twitter-like service Weibo, several users wrote that the news readers have “no soul.” One user joked: “What if we had an AI leader (of China)?”

Others wondered about the effects robots might have on employment and workers. “It all comes down to production costs, which will determine if [we] lose jobs,” one Weibo user wrote.

This photo illustration shows a man watching an artificial intelligence (AI) news anchor from a state-controlled news broadcaster, on his computer in Beijing, Nov. 9, 2018.
This photo illustration shows a man watching an artificial intelligence (AI) news anchor from a state-controlled news broadcaster, on his computer in Beijing, Nov. 9, 2018.

Some people argued that only low-level jobs requiring heavy labor will be easily replaced by robots. Others praised the technology as a way for companies to make money from low-cost labor machines.

A writer for the MIT Technology Review declared Xinhua’s news reader announcement as “fake news.” China’s news presenter “isn’t intelligent at all,” the report said.

MIT Technology Review said the presenter itself is little more than “a digital puppet that reads.” It added, “The ‘AI’ in this case is the software that learns what makes a convincing-looking face and voice.” While the demonstration was clearly evidence of improved technology, it was “a very narrow example of machine learning,” the report said.

Some businesses have experimented with similar technology for possible use in news operations.

Britain’s BBC recently released a video that used machine learning to make it look like one of its news readers was speaking different languages. The London-based company that developed the system says its goal is to “remove the language barrier” for many different kinds of video across the internet.

Also, developers working on Amazon’s Alexa smart speaker system have worked to create a special news reader voice for the electronic assistant.

Trevor Wood works on text-to-speech technology for the Alexa system. He said in an internet post that a machine learning system was able to help Alexa ‘learn’ to speak in a news-like voice after only a few hours of examining data.

Such a system could be further developed to provide other kinds of speaking voices for different situations, Wood said. Amazon is reportedly preparing to launch Alexa’s news reader voice in the near future.

I’m Bryan Lynn.

Bryan Lynn wrote this story for VOA Learning English. His report was based on reports from the Reuters news agency, VOA News and online sources. George Grow was the editor.

How do you feel about robotic news readers? Would you trust the news delivered by one? Write to us in the Comments section, and visit our Facebook page.

-————————————————————————————————————————————

Quiz - Future of News: Will Robot Readers Take Over?

Quiz - Future of News: Will Robot Readers Take Over?

Start the Quiz to find out

________________________________________________________________

Words in This Story

artificial intelligence n. ability of a machine to use and analyze data in an attempt to reproduce human behavior

data analysis n. process of using computers to gather and inspect facts and other information

soul n. the part of a person that is not their body, which some believe continues to exist after death

determine v. discover the facts or truth about something

fake adj. false, not true

puppet n. a person controlled by someone else

software n. programs used to make a computer do different things

convincing adj. able to make you believe that something is true

XS
SM
MD
LG