The world around us is changing, with technology right at the center, driving the change. One piece of tech that is currently making the rave is deepfakes. In 2019, a video that featured Mark Zuckerberg claiming he has complete control over the stolen data of billions of people started making rounds online. Likewise, another video of Morgan Freeman questioning reality. Although both videos turned out fake – AI deepfakes, it shows how far we’ve come.
What is Deepfake Technology?
Deepfake is a combination of two words – deep learning and fake. It is a fusion of artificial intelligence and machine learning and uses Generative Adversarial Network (GAN) architecture and AI Neural networks to reconstruct inputs from a simpler representation. In other words, two AI algorithms battle each other. One GAN creates a fake image, while the other tries to detect if the image is manipulated or not. They both repeat the process until the latter cannot differentiate between false and real information.
Deepfake technology uses images, audio clips, and videos to create fake videos, audio, or images that look real. As a result, they often feature individuals in situations that never happened. The technology can even create a replica of original content and make it realistic.
How are Deepfakes Used?
Deepfake is not the same as video or photo editing. Instead, it is more about manipulation. Creators can swap the face of an individual with another, as is common in AI-generated porn. The generators may overlay real faces onto the bodies of adult performers or create an entirely new figure via prompts. The prompt converts texts into pictures or videos, so you can create porn using tags like skin, breasts, and face.
Beyond swapping faces, deepfake technology is capable of lip synchronization. In other words, it takes a video of a man or woman speaking and changes the lip movement to match another audio track. Lastly, the tech can swap an entire body with another one. In a nutshell, deepfake is capable of highjacking an individual’s face, body, and speech to suit the creator’s narrative.
The Rise of Deepfakes
Deepfake first gained the world’s attention in the late 2010s, although its academic study dates back to the 1990s. Initially, deepfake was commonly used for low-level parody and entertainment. However, it soon became notorious, and the public started citing its limitless use and threat. Ian Goodfellow and others developed GANs in 2014, and they served as the foundation for deepfakes. Apart from GANs, open-sourced projects and applications facilitated easy access to producing and distributing deepfake content online.
Improvements in artificial intelligence gave deepfakes more proficiency. They became adept at analyzing and replicating human features, expressions, and voice patterns. Plus, they can manipulate videos and audio in real-time. Consequently, it gave rise to live deepfake apps, which opened the doors to limitless opportunities.
Deepfakes with AI – The Good and the Bad
Deepfakes or synthetic media have far-reaching consequences that cut across various aspects of our lives. According to statistics, nearly 90% of online video content could be synthetic by 2030.
Deepfakes are transforming the entertainment industry. They allow filmmakers to resurrect deceased actors or create digital stunt replicas. One such is the series of hyper-realistic deepfakes of Tom Cruise that blew up on TikTok in 2021. The technology reduces cost and for adult content, allows individuals to fulfill their fetishes.
Deepfake technology creates lifelike simulations that are handy for education and training. For instance, in medical simulation or language learning. Businesses employ it in marketing to create personalized campaigns, and artists leverage deepfakes to explore different concepts. On the downside, the potential for misinformation is enormous.
Unscrupulous entities utilize deepfakes to spread misinformation, defraud individuals, and manipulate public opinion. Celebrities, political figures, and ordinary individuals are all targets. Apart from this, creating the likeness of an individual as synthetic media without permission constitutes a violation of privacy rights, which may have legal consequences. As the tech grows, there are concerns about potential misuse by state-sponsored actors and individuals, which has led to calls for increased research into countermeasures and detection measures.
Bottom Line
Deepfakes, like other technology, can be good or bad. It all depends on the context of usage. The blurring lines between fiction and reality raise serious ethical concerns about consent and authenticity. That said, its responsible use presents endless opportunities for the user.
An example of a deepfake porn video is Marilyn Chambers going down on Hayden Winters.