Would You Get Your News From An Artificial Intelligence Newsreader?
Riddle me this: if the robots do wage war, will the artificially intelligent newsreaders tell us?
It's a question for perhaps the not-so-distant future, as China's state news agency unveils its first virtual news anchors.
Xinhua News and Chinese search engine Sogou debuted their "composite anchors", which are a combination of digitally-manipulated faces of real newsreaders and computer-generated voices, at the World Internet Conference on Wednesday.
“Hello, everyone. I’m an English artificial intelligence anchor. This is my very first day in Xinhua news agency," Xinhua's newest AI employee said in his introductory video at the conference.
The anchor's machine learning programme is able to synthesise realistic-looking speech, lip movements and facial expressions, Xinhua said, to deliver news with "the same effect" as a human.
Potentially able to work 24 hours a day -- provided there is an editor there to feed through text -- the new technology may help to cut down on production costs and deliver breaking news in a more timely manner.
“AI anchors have officially become members of the Xinhua News Agency reporting team,” Xinhua said. “They will work with other anchors to bring you authoritative, timely and accurate news information in both Chinese and English.”
According to the Chinese broadcaster, the presenters "can read texts as naturally as a professional news anchor".
Not only is this highly debatable (see noticeable robotic reading in the video above), but it begs the question: would the public accept an artificial figure reading their nightly news?
"This is scary stuff, and arguably I am staring directly at my future replacement," veteran 10 News First newsreader Sandra Sully said of the technology.
"Honestly, it's much better and more impressive than I expected, however there is no natural cadence in his delivery and that contributes to a lack of authenticity. Consumers are already so wary of fake news and a robotic version only enhances that level of mistrust."
A Fake Newsreader In The Fake News Era?
The impact of this AI technology has already been felt in the media landscape.
"It’s the same technology that’s going to be used to make politicians say things that they never said," Professor of artificial intelligence Toby Walsh told 10 daily.
In April this year, BuzzFeed created an eerily convincing video of former US President Barack Obama appearing to speak against Donald Trump. Known as 'deepfakes', these videos created with AI technology are deeply concerning for several reasons, including their ability to impact elections and leave viewers at a loss when deciding what's real and what isn't.
"We’ve already had President Trump say things are fake news when it’s been caught on camera and we no longer can trust what we see with our eyes unless you were there in person you cannot trust that it actually happened anymore," Walsh said.
In his latest book, Professor Walsh predicted the evening news would be produced without any human input, from the newsreader to the writing of reports.
But in the face of this possible reality, he admits the public may not buy into it wholeheartedly.
"We quite value the relationship we have with the news readers that we get to know over many years," he said.
"We’ll probably keep that although you can see the financial advantages to replacing those very expensive newsreaders with AI robots that never sleep and never require wages."
Walsh said China is quickly starting to lead the way in AI development, with the nation investing in the field "in a big way" as they seek economic and military dominance.
Elsewhere, the technology is making its presence felt across several professions, from medicine to transport. AI has even given human creativity a crack, with a portrait of Edmond de Belamy painted using AI selling for $610,000 at auction in October.
Feature Image: Xinhua News
Contact the author: firstname.lastname@example.org