New High Tech Artificial Intelligence Makes Fake Look Real

The Truth Behind Fiction


In the dawning era of advanced photo and video editing, discerning what’s fake or real is no simple task. Now with the invention of new voice tech that can simulate any human voice on video, we are further at risk of being fed false information through the internet and social media. If fake news was scary in the past, it is now a nightmare. When these advanced technologies land in the wrong hands, it can disrupt all our lives.


As if being misunderstood in written language were not a big enough problem, now you can be misunderstood in spoken language too. Now artificial intelligence can mimic a person’s voice matched to an unrelated video and make it look like you said or intended something you didn’t. As technology provides a means to put words in your mouth, you will be judged and held accountable for things you didn’t even say. Damage control will be impossible once someone spreads fake videos about you. Even if one issues a correction later, it will not fix your already ruined reputation. What implications does it have for the legal system when people submit fake videos as “evidence?” Will government officials, leaders, and your average citizen be able to discern the truth?


Artificial Intelligence


Artificial Intelligence has become a sophisticated technology that is now able to learn the mannerisms and decode the speech of any individual to produce the most natural sounding voice. With it, you can literally mimic anyone in a natural sounding way while taking into account their intonation, inflections, pacing and voice.


Adobe’s project Voco allows you to photoshop out sounds and alter the recordings of anyone’s speech. If enough of a person’s voice has been recorded and studied by it, a simple cut and paste function can edit out the person’s voice and replace it with a natural sounding one. If you have enough data in your library to imitate the person, then anyone can technically appear to say anything.


Adobe’s project results are terrifying. But the next step of pairing simulated audio with doctored videos is more terrifying. Now somebody with malicious intent can make you do and say anything they want.


Don’t believe that the fakeness can look so real?


Watch below for a sample video of Barack Obama speaking. The Researchers at University of Washington produced this footage based on a deep learning algorithm of the subject’s speech and mannerism. The video of Barack Obama giving a speech is completely bogus.


Video using New Artificial Intelligence Technology


Although the creative use of such impressive voice technology on video has broad powerful potential in the entertainment or internet industry, when the wrong people use it in the practical world, its consequences can be catastrophic.


With the advancement of this high tech artificial intelligence, truth and fiction are impossible to tease apart. We live in a world where fake news is rising, and what you see and hear cannot be trusted.


Read more about How AI Can Mimic Any Human Voice and the Impact of Fake News




Leave a Reply