According to the Massachusetts Institute of Technology, virtual broadcasters use China's AI as fake goods

In November 2018, many people were amazed when China's news agency Xinhua launched two power-operated virtual television broadcasters. According to Xinhua, the virtual announcer is programmed to read voice-based news synthesized by computer. A video recording of the first session of this virtual announcer was also announced by the company.

However, according to a reporter from MIT Technology Review, MIT Institute of Technology, this is not an artificial intelligence announcer but just "a voice-reading puppet". This is a computer-generated product with a slightly stiff voice, slightly different lips but not a 'smart' product.

According to the Massachusetts Institute of Technology, virtual broadcasters use China's AI as fake goods Picture 1

Sogou, the father of these virtual broadcasters, used a software to create looks and voices that sounded like a real announcer.

But according to researchers at MIT, this is an impressive product that can be called "artificial intelligence announcer" but is only a very shallow example for "machine learning".

And in fact, anyone can create such videos by deepfakes, a program that allows users to use someone else's face to replace someone's face in the video. This technology brings many benefits in the future such as improving animation ability, computer skills and game development technology. However, the fact that a computer has the ability to integrate familiar, famous faces into any human body also contains worries. One of them is the appearance of very indecent videos using this face transplant technique on the Internet.