The deepfake Putin told Putin that he was a “student from St. Petersburg” and asked him about the rumored doubles he seems to have. Vladimir Putin replies, “You can be like me and speak in my voice. But I thought and decided that only one person can be like me and speak in my voice, and that person is me. By the way, this is my first double.”
The deepfake of Putin introduced himself as a “student from St. Petersburg” and asked how many doubles Putin has
“You can be like me and speak in my voice. But I thought and decided that only one person can be like me and speak in my voice, and that person is me. By the way,… pic.twitter.com/YNq5a8ZhMC
— NEXTA (@nexta_tv) December 14, 2023
Deepfakes come from “deep learning,” or machine learning. An AI model is fed audio clips of the person talking, and it “learns” how the individual speech patterns sound, the inflection of their voice, the tone, and other identifiers. The AI model begins to create a synthetic recording of the voice. The same machine learning is used with body images that learn a specific person’s ways and unique movements and generates the likenesses from there.
One of the great deepfakes this year was David Beckham, who fluently speaks nine different languages — he only speaks one. The Beckham video set up many online questions and gave Beckham a platform for some TikToks. From the Beckham incident and other examples of AI methods — many companies have determined to use deepfakes of celebrities with or without their permission.
A digital media and marketing professor at Carnegie Mellon University’s Heinz College of Information Systems and Public Policy may have said it best, “We’re having a hard enough time with fake information. Now we have deepfakes, which look ever more convincing.”
Featured Image Credit: Andrea Piacquadio; Pexels