The essential deepfakes, these increasingly impressive digital effects, are flooding social networks. Fake videos, often funny and sometimes wrongly taken seriously by internet users.
Hundreds of thousands of TikTok users recently watched this video by Tom Cruise on TikTok. The “Mission Impossible” star appears there in close-up, talks to the camera and then plays on a golf course. A priori, nothing surprising except that it is an ultra-realistic special effect performed by the @deeptomcruise account. Most of the users saw it fire, believing it was the real Hollywood actor, and announced its arrival on the favorite youth network. On another, he performs a magic trick with a coin.
I love magic!
♬ Original sound – Tom @deeptomcruise
♬ Original sound – Tom
Tom Cruise, bigger than life? “The resemblance was striking, it was stunning.” “Almost perfect, almost impossible to spot the ruse!” “Amazing! How did you do it?” There have been a lot of reactions on social networks in the past few days as this technology raises several questions.
The members of the royal family and the portrait of the Mona Lisa have also recently been noticed on the internet thanks to an artificial intelligence technology that can bring a photo to life. In large digital companies, the race for artificial intelligence is in full swing to simulate reality as well as possible.
Can these virtual clones be dangerous, especially if the tools are placed in malicious hands? Shall we ban them? For Chris Umé, VFX (‘visual effects’ translation) specialist who created Tom Cruise’s videos, sooner or later a law will be needed to regulate this technology. “We need to think about a system to detect deepfakes and possibly flag them,” he told the newspaper 20 Minuten. I think journalists have an important role to play in confirming the sources by saying that this is true. When Photoshop first came out, people didn’t realize you could edit images. Now everyone knows that. “
This complex digital technology, which is becoming more widespread and more sophisticated, makes it possible to reproduce a person’s face on a moving body. Produced with highly developed software when it is close to perfect, it is also available to everyone thanks to free applications on Instagram, SnapChat and others.
French Faker, the reference from Toulouse
The 29 year old Toulousain French Faker is a professional in this field. He works regularly with the program “C’est Canteloup!” Together that is now based entirely on this effect. He also “replaced” an actress with an understudy during an episode of “Plus belle la vie” last November. A first in France.
Also read: VIDEO. The special effects of French Faker, this Toulouseain who works for Canteloup and “Plus belle la vie”
His latest video has already received more than 300,000 views. It shows a false speech by Emmanuel Macron from the short film “The Underworld of Power”. The President of the Republic is giving a crazy speech against the backdrop of a nuclear attack and latent eroticism. An unsettling realism that will be reinforced with the next technological advances, proposing a 2022 presidential campaign riddled with deep fakes. “The next step for deepfakes will be to be able to mimic voices more closely,” said Yanis, aka French Faker, in an interview with L’Express a few days ago.
The defense of Dieudonné
When it becomes difficult to separate the true from the false, one doubts all images. An unexpected defense strategy for controversial comedian Dieudonné attempted in Versailles on Monday March 15 for contempt for a judge. In the incriminated video, which was published on YouTube on April 8, 2020 and has since been withdrawn, we hear her severely criticizing the demands of a judge from Nanterre against a pharmacist and, in particular, comparing the judge with the women who worked together in the Nazi regime . “It’s not me in the video,” he told the audience.
“In this case, Dieudonné was the victim of a fake video that was supposed to harm him,” said his lawyer Me De Stefano, asking for his release and saying that his client was also affected by “five other cases” of “videos deepfake” be. in Paris. “It is time for the press to make their 2.0 revolution because today the deepfake image is no longer evidence,” he added.
Difficult to recognize, even by machines, the deep panic more and more the ghosts.