For decades, children have been told not to believe everything they see on the internet. But now we need to extend this advice to adults.
There are more and more incidents in which artificial intelligence (AI) technology is used to manipulate video and audio in a way that mimics real life, referred to as “deepfakes”. And now, the world’s first “certified” deepfake video has been released by AI studio Revel.ai to help set an example of transparency.
In this video, Nina Schick, a professional AI consultant, is seen warning about how “the lines between fact and fiction are blurring.” However, as you can guess from the beginning of the article, the video seen in the video is not actually the real Schick and the video is cryptographically signed by digital authenticity company Truepic and declares that it contains artificial intelligence generated content.
“Some people say the truth is a reflection of our reality,” says the “fake Schick” produced by deepfake technology, slowly and clearly. The explanations continue: “We are used to identifying it with our own senses. But what if our reality is changing? What if we can no longer rely on our senses to determine the reality of what we see here and here? We are at the dawn of artificial intelligence and the lines between fact and fiction are already blurring. It’s a world where shadows are mixed with reality, and sometimes we need to radically change our perspective to see things as they really are. This deepfake was created by Revel.ai with permission from Nina Schick and is cryptographically signed by Truepic.”
- What is Deepfake? How does deepfake technology work?
Appeared with porn videos
Deepfakes are forms of artificial intelligence that use “deep learning” to manipulate audio, image or video, creating hyper-realistic yet fake media content. The term was coined in 2017 when a Reddit user posted manipulated porn videos on the forum.
The videos had the faces of celebrities like Gal Gadot, Taylor Swift and Scarlett Johansson swapped with pornstars without their permission. Another dangerous example of deepfake emerged last year when Ukrainian president Volodymyr Zelensky appeared to have surrendered to Russia in a video that was widely circulated on Russian social media. In this fake video, the president was urging the Ukrainian soldiers to lay down their arms and bow to Putin’s invading forces.
At first glance, this video looked realistic enough, though mindful internet users quickly found inconsistencies such as the color of Zelensky’s neck and face, his odd accent, and the pixelation around his head.
While deepfake videos may have an entertainment value, some experts have warned for years about the dangers they can pose.
Many experts point out that these deepfake videos can go beyond reputational damage and create serious security hazards that can go up to the national level and can be used for their own purposes by “enemy factors”.
With the rise of free text-to-image and text-to-video AI tools like DALL-E and Meta’s “Make-A-Video”, manipulated media is likely to become more commonplace. In fact, it is estimated that by 2025, 90 percent of online content will be produced or created using artificial intelligence. In short, we may now have to be much more skeptical of everything we see on the internet.
49703
