The dangers of deepfakes have become unprecedented. They have departed from just mere look-alikes that seem a bit off to shocking but scary good videos that tend to go viral. The human race has reached a state where the deepfakes are incredibly well done and very hard to identify.
The technology responsible for deep fakes is getting more efficient and online applications are making it very simple. With the eventual enhancement, it becomes challenging to differentiate between reality and fakeness. So one might wonder; what is the concept behind deep fakes? And what steps can be taken that make it easier to spot them? And does human civilization need to be aware of this technology as it could lead to something eerie?
What are Deepfakes?
If we break down the word, “fake” in deep fake is pretty obvious whereas “deep” indicates deep learning by a machine. This is attributed to artificial intelligence. So the purpose behind deepfake is to impersonate people and use their faces to proclaim things non-consenually.
Working Process of Deepfakes
So here is how it goes; you need to take a video of yourself and a video or photo of a celebrity that you intend to create an ai deepfake of. After that, you can use a free software called DeepFaceLab.
In the first step, convert your video into hundreds of images which are called frames. Then in the next step separate the faces from the images. The software will pin your eyes and mouth to the center of the frame so that it can be aligned with the target face. After that, we have step three which is the machine learning phase followed by adjustment of the skin texture and then amalgamating the two faces on the end picture.
Applications
The process mentioned above might take a couple of days but there is no denying the fact that creating deep fakes is getting much easier as the technology becomes more efficient and swift. Now the question is how can we utilize this technology?
Deepfakes have many applications that are often considered creative and playful. For instance, David Beckham ran an anti-malaria campaign in which he was speaking in nine languages. This advertisement utilized celeb deepfake technology to make that happen. This is how it could be used for a positive purpose.
However, there are instances when deepfake technology has been used to harm people. For instance, recently a woman in South Korea was tricked into giving away $50,000 to an unknown person pretending to be Elon Musk. Various news sources and local media state that the woman was deceived by a deepfake video in which Elon Musk the CEO of Tesla was shown. The person made the impersonation so convincing by showcasing all the details of Elon’s life. However, he talked mainly about his children and family which developed a level of trust between them. The scammer then manipulated and emotionally blackmailed the lady into investing in a fraudulent activity. As a result, she lost all of her life savings.
on the other hand, there are other cases where deepfakes have blurred the lines between privacy and verification. A website called “MyHeritage” uses pictures of your deceased relatives. For some people, it would be fine but in reality, it puts people in a very precarious position.
Biggest Problems
The biggest problem with Deepfakes is that they are made without the consent of people. Celebrities and women are the biggest victims of this technology. There are many people out there who are using the photos of their celebrities and even normal working women and then deep-faking them into pornographic actors. According to a recent study that was conducted in the Netherlands, 95% of the celeb deepfakes were coerced porn. It shows us that deep fake can turn into something very nasty. Experts claim that it has deep impacts on the mental health of people. It often leads people towards depression and eventual suicidal tendencies. For instance, a woman in Pennsylvania allegedly utilized this technology to discredit her daughter’s rivals in high school by creating fake videos of them in explicit ways.
The Potential Danger
Deep fakes have blurred the lines between what’s real and what’s not. It is becoming very hard to establish trust with someone, especially in the digital space. Every piece of information that you may see may be aimed at manipulating you. The recent advent of deepfake applications has turned photos and videos into a cheap way of spreading misinformation. AI has established the ground base for a 5th generation of warfare. Now one does not need to threaten a world leader or a nation vis-a-vis hard power, all you need is the ability to create a realistic fake video. Therefore, deepfake detection is important.
Conclusion
Conclusively, one needs to understand that deepfakes create a scenario where reality can be mistaken as fake and vice versa. Hence, one needs to learn how to spot celebrity deepfake, for instance, by becoming aware and not being fooled by everything that you may see online.
