There was a time when it was said that pictures did not lie. But PhotoShop ruined that. Video editing made it worse. And just when you thought it couldn’t get any worse, a team of 3 scientists led by Google computer scientist named Supasorn Suwajanakorn has invented a lip-reading technology that can learn how anyone speaks and then use the data to create convincing, fake videos of them saying anything he wants.
Is that sinking in? Do you have any idea how scary that is? This is video editing taken to nightmare levels. Anyone with access to this technology can make a video of you making a terrorist or treasonable statement and put you in big trouble.
This new lip-reading technology watches videos of a person talking to learn how to imitate the way they talk. It then can be used to create a convincing, fake video of them. Here is an example of a synthesized video of Barack Obama.
The result is realistic and convincing. A fake video. The results were so convincing that Google hired Supasorn. Again, this is a nightmare. Photos now lie, and even videos can now lie in the most convincing way.
One question I have on my mind though: when working on this new technology, what good did the creators have in mind that this would bring to the world? Wipe out hunger? Kill poverty? Or is this one of those things that was created because we can?
Founder of MobilityArena. Yomi’s journey in mobile started in 2001. Besides obsessing over mobile phones, he also started creating WAP sites (early mobile-friendly websites created with WML). He began writing about phones in 2004 and has been at it since then. He has owned over 200 devices, from Symbian, Palm, PocketPC/Windows Mobile, BlackBerry/BB10, webOS, Windows Phone, Firefox, Ubuntu Touch, to Android, iOS, and KaiOS operating systems.