Home | Science | Technology | AI Fake Video Technology Poses New Threat to Public Trust in Media

AI Fake Video Technology Poses New Threat to Public Trust in Media

1021 Views

By Staff Reporter

May 3, 2018
12:32 pm
Los Angeles, CA

At the TED conference on April 11th this year, a new Google engineer, Supasorn Suwajanakorn, demonstrated the video production algorithm developed by him and his associates to produce realistic fake films. Even though it is untrue, it has become so “real” that it is difficult to distinguish from the naked eye. Supasorn Suwajanakorn used his own fake Obama video as an example. After inputting a 14-hour Obama video lecture, he can use an algorithm to analyze Obama’s mouth shape and subtle expression changes. He can then create a 3D dynamic model. Then just enter any piece of audio. The 3D model will follow the same words and form a fake film. He publicly stated in his talk that such A.I. “fake video” technology may bring unpredictable danger. In fact,

Supasorn Suwajanakorn used his own fake Obama video as an example. After inputting a 14-hour Obama video lecture, he can use an algorithm to analyze Obama’s mouth shape and subtle expression changes. He can then create a 3D dynamic model. Then just enter any piece of audio. The 3D model will follow the same words and form a fake film.

He publicly stated in his talk that such A.I. “fake video” technology may bring unpredictable danger. In fact, this unpredictable danger is the last line of defense that will undermine the trust between people. It may eventually lead to irreparable chaos.

The truth of information.

This refutation of false material is still based on the rebuttal message or video being secure, personally published, and not being subject to alteration or fabrication by another person.

But, when the above A.I. “pseudo-film” has become so popular that everyone can use it, people with a particular intent can even create a true “fake film” of state leaders. The content may include announcing operations against another country, or showing false news that is harmful to national policies, or to announce fake information that causes turmoil in the exchange market, and then use social media to give it mass retransmission to the public, and it is very possible that before the real message is clarified, it could have caused huge damage or loss around the world.

The same situation can affect anyone. If a person intentionally distributes another person’s “fake video” to their relatives, friends, colleagues, superiors, or competitors, it is very likely that they will be able to produce irreversible effects.

And all this chaos is easy for anyone to create: A.I. tools work on home computers and even mobile phones.

When such chaos continues to evolve, no-one can believe in social media, or anything in regular media and mainstream media, or they have to expend even greater costs and effort to clarify or verify all media content. Trust between people will collapse overnight. Unless a person painstakingly clarifies or communicates with others in person, any video information that is distributed on electronic media may be falsified or manipulated. In the end, everyone will be in danger, and no one can return to the peace of mind they had before.