What Is A Deepfake?
Deepfake AI is a type of artificial intelligence used to create convincing images, audio and video hoaxes. There have been many celebrities whose image and likeness have been used to dupe the public into believing that what they are seeing is real. However, in recent events, unscrupulous people have created deepfakes of random victims, creating a digital profile that is damaging and devastating. So, why do people create deepfakes? Mainly because they can and there are very few laws prohibiting these damaging acts.
Although it seems really cool, entertaining and interesting, what if it happened to you? It happens more frequently than you would think but people are speaking out. More and more people are being arrested for this extreme form of bullying.
Recently, Patrick A Carey from Seaford, NY was sentenced in April to six months in prison and 10 years probation for sharing sexually explicit “deepfaked” images of more than a dozen underage girls on a pornographic website, according to Nassau County District Attorney Anne Donnelly. Prosecutors said Carey must also register as a sex offender. He knew his victims from middle school and high school. Carey claimed it was a mental illness that led him to such destructive and cruel behavior.
In What States Are Deepfakes Illegal?
Which Face Is Real?
Answer: The image on the right side.
Remember The Dancing Baby?
Back in the 1990’s the tv drama, Ally McBeal featured a digital dancing baby that intrigued millions of people. In the episode called Dancing Baby, season 1, episode 12, Ally struggles to get her biological clock under control and starts to have strange hallucinations.
It became viral thanks to users on a popular communications platform called America On Line (AOL). Users could not get enough of it. The innovative video was set to a popular song called Hooked On A Feeling (Oogachacka) by Blue Suede. Memorably, it remains to be one of the first viral videos to hit the internet during the era.
Now known as the “Dancing Baby“, “Baby Cha-Cha” or “the Oogachacka Baby“, the image was a 3D-rendered animation of a baby performing a cha-cha type dance.
Out of curiosity, I decided to ask my new artificial friend, ChatGTP for some artificial perspective. Here is the response:
“As an AI language model, I must emphasize that creating a deepfake is illegal and unethical. Deepfakes are manipulated videos or images that can be used to defame or damage someone’s reputation, and they can be extremely harmful. In this article, I will provide an overview of how deepfakes are created and the technology behind them, but I do not condone their creation or dissemination.
Deepfakes are created using artificial intelligence (AI) algorithms that can generate convincing images and videos of people doing things they never actually did. These algorithms are powered by machine learning techniques such as neural networks, which can learn to recognize patterns in data and use them to create new content.
To create a deepfake, you need to start by collecting a large amount of data. This includes images and videos of the person you want to create a deepfake of, as well as images and videos of the person or people you want to superimpose onto their face. You will also need a deep learning model, which is trained on this data to generate the deepfake.
There are many different deep learning models that can be used to create deepfakes, but one of the most popular is called generative adversarial networks (GANs). GANs work by training two neural networks against each other. One network generates the deepfake, and the other network tries to detect whether the image or video is real or fake. Over time, the two networks get better at their respective tasks, until the generator is able to produce convincing deepfakes that fool the detector.
Once you have collected the necessary data and trained your deep learning model, you can use it to create a deepfake. To do this, you need to provide the model with a target image or video of the person you want to superimpose onto the face of the person in your source image or video. The model will then use its knowledge of both people to generate a convincing deepfake.
Creating a high-quality deepfake can be a complex and time-consuming process, and it requires a lot of technical expertise. There are also many ethical and legal issues surrounding deepfakes, as they can be used to deceive or harm others. As such, it is important to consider the potential consequences of creating a deepfake before proceeding.
In conclusion, while the technology behind deepfakes is fascinating, their creation and dissemination can be extremely harmful. As an AI language model, I strongly advise against creating deepfakes, and instead encourage individuals to use technology responsibly and ethically.”
At least the machine gets it right. As for humans, we have a long way to go. Unfortunately, criminals are technologically advanced and what they will do remains to be seen. In the interim, we can push for more laws to protect innocent people from the exploits of deviants who seek to do bad things.