Take a look at the video in this post.
It is of the latest news announcer to be unveiled by China's state news agency, Xinhua News. He looks young, serious and smart in his sharp suit and tie, just like many other TV announcers all over the world. What marks him out though is that he isn't real.
He's a digital composite created from footage of a real host and is powered by artificial intelligence.
According to reports, an artificial intelligence system has been used to synthesize the virtual presenter's lip movements, expressions and voice which are based on real Xinhua presenters.
Xinhua plans to use its stable of virtual presenters on its websites and social media channels which will reduce running costs. The anchors will keep viewers informed as texts are typed into their system. Currently, there are two virtual anchors, an English-speaking version and a Chinese-speaking one.
"I look forward to bringing you the brand-new news experiences," said the English-speaking AI anchor in an introductory video.
Although the anchor doesn’t look completely natural, many commentators have said that this is a very good first attempt. Among the criticisms are that the synthesized voice is flat and lacks the pace, rhythm and emphasis that you get with human presenters.
The range of facial expressions also appears to be limited. But with the way advances are rapidly coming in machine learning it may not be too long before improvements give us AI anchors that are indistinguishable from the real thing.