A deep fake is a piece of visual media that generates someone else’s face on another person’s body. Usually reserved for videos, deep fakes are known for their astoundingly realistic results that make it difficult to discern between original media and artificial generations.
Whether you have been amazed by the TikToks of Tom Cruise dating Paris Hilton or concerned by President Obama laying out uncharacteristic statements, you must have heard about deep fake technology and its ability to create realistic videos by now.
While the results from deep fake are nothing but amazing, they also bring about a variety of questions about its ethical and societal impact. But before you get into exploring those topics, you have to understand exactly what is deep fake and how does it create realistic videos that make you look twice.
To help you through this learning curve, here’s a lowdown on deep fake meaning, its technology, and its overall impact on the present and future of media.
A combination of the words “deep learning” and “fake,” the term “deep fake” is a direct reference to artificial intelligence (AI) and forgery, respectively. This makes the media’s intentions known from the get go at how it is generating these unbelievably realistic videos while also giving them the distinction of being mere imitations of the people they are depicting.
With that being said, not every deep fake that is generated and distributed online has the disclosure of being a deep fake in the first place. This makes it crucial for every internet user to learn deep fake meaning as well as its overall impact on the media that we consume.
A deep fake is made with the help of AI and manual manipulation of visual media. Through AI, the creators of deep fakes are able to train the image or video generation system on hundreds or thousands of images of the person that they are going to integrate in the deep fake. Through manual manipulation, they are able to adjust contrasting issues such as video quality and face edges to make the video seem as real as possible.
This means that creating a deep fake is not an easy feat. However, many online deep fake generators allow you to make your own deep fake within a few clicks, especially when it comes to celebrities. It is because these systems have been trained with the hundreds and thousands of images of celebrities that are needed to make impressive deep fakes.
In turn, even after learning what is deep fake, making your own deep fake of a non-celebrity can be difficult. While you can use online deep fake generators, you can often tell them apart from real videos and images.
The exploration of deep fake meaning has blown up in popularity only recently. But the first deep fake was made way before the trend started to take TikTok by storm. In fact, the first popular iteration of deep fake was made possible all the way back in 1997, when the software Video Rewrite made its presence known.
Through Video Rewrite, users could generate deep fakes of an already existing video, with the movements of the person’s mouth adjusting to a different audio. However, the deep fakes that were generated through Video Rewrite were easily identifiable. This led those who were interested in the technology to continue developing programs that could create more advanced and realistic-looking deep fakes.
Deep fake has come a long way since its introduction in 1997, with videos being able to depict the subject’s face on another person’s body almost seamlessly. When the person in the video speaks, the facial movements of the subject’s face adjust to it by using the images that the system has been trained on beforehand. This also goes a long way towards integrating the relevant expressions that go along with certain sentences.
While many deep fake generators still falter at putting a seamless outline of the subject’s face with the body of the person in the image or video, certain programs create highly realistic videos that make it difficult to tell a deep fake apart from real videos. The Tom Cruise videos that went viral on TikTok are an excellent example of the latter.
Created by AI and VFX artist Chris Umé, the Tom Cruise videos show the Top Gun star in everyday situations like a typical influencer. You can see Cruise shopping, riding around in a golf cart, and even telling Paris Hilton that he cannot believe that they are a couple. From the visual finesse to the sound effects, everything makes you believe that you are looking at the real Tom Cruise in these videos.
After learning what is deep fake, it will be easier for you to understand how Umé uses AI to mimic Cruise’s facial expressions in the videos. With thousands of images fetched from the Mission: Impossible star’s illustrious career as well as his public appearances, Umé uses AI to adjust the star’s facial expressions to an impersonator, Miles Fisher.
While the AI does the job of generating highly realistic succession of images, the Cruise impersonator speaks in his voice and emulates the actor’s mannerisms. When put together, it creates an eerily realistic deep fake that is almost impossible to tell apart from any real appearances that the Jerry Maguire star has made over the years.
Essentially, this turns the deep fake meaning over its head. What was once a way to manipulate the speech in existing videos has now become an art form to create previously unimaginable content. After all, a world where Tom Cruise is dancing to a viral Lady Gaga song is much better than the one without, isn’t it?
But as entertaining and positively enthralling deep fake videos can be in this context, they also pose grave ethical concerns for the content that they can generate. That is where you need to look at what is deep fake able to achieve when it falls in the hands of bad actors.
Similar to any piece of technology or basically any type of invention, deep fake has its pros and cons. While the technology opens doors to create media that was not possible before, it also creates opportunities for harmful parties to develop images and videos that can leave lasting harm on the people and groups involved.
Since deep fake can create content out of practically nowhere, it is a highly dangerous tool that can be used to spread harmful imagery as well as misinformation. This can affect people on a personal and societal level, while also turning large groups against certain causes or political issues. In its more extreme form, deep fakes can also sow the seeds of doubt or violence in people’s minds.
Over the past few years, deep fake technology has been used to create false images and videos that depict the subject in situations that can be harmful to them. This includes instances such as fake pornography where the subject’s face is put on top of the body of an actor. Other situations involve putting the subject’s face on inciting statements that may lead people to believe false accounts and instances.
Through deep fake technology, malicious parties can generate content including but not limited to:
- Sexual abuse media
- Pornographic media
- Fake news
This use of deep fake has led people to perceive it as a harmful technology that needs to be controlled or stopped in its tracks. Nonetheless, advocates of deep fake such as Umé who generates content for the Tom Cruise deep fake account on TikTok believe in the positive contributions of the technology. This includes use cases such as creating videos using actors or figures from yesteryears.
After learning what is deep fake and what does it do, you can better compare the advantages and drawbacks of the technology across a number of aspects. The technology may likely see widespread adoption in the near future. But if and when that happens, you can make an informed decision about your participation in the spread and usage of this technology.
With deep fakes, marketers can create engaging adverts using archival footage or historical figures. This creates captivating outreach campaigns with images and videos.
Deep fakes can also allow creators to generate new facial expressions for the same person in the video. At the same time, they can also replace the original audio with new dialogue.
Deep fake can be a fresh form of art that can allow AI and VFX artists to become experts in a whole new field. This can usher in an exciting era of content production and refinement.
When TV shows and movies need to create flashback scenes, they usually have to look for younger actors who may or may not look like their counterparts. With deep fake, creators can deage existing actors for easier production and better visual impact.
Deep fakes can also combine the facial features of people other than the subject and the person in the original media. This can cut costs in terms of makeup and VFX, while also reducing larger post-production work.
Deep fake can put people in imagery that can be extremely harmful. Some of these examples include fake pornography and sexual abuse content. This paints deep fake in an extremely bad light.
Deep fake technology can also be used to spread misinformation by depicting famous figures, especially politicians, delivering fabricated statements. This makes the technology dangerous for society.
The content generated through deep fake can be used to incite violence and spark anger among certain groups. This can lead to social harm and safety issues on a large scale.
By depicting famous businessmen or celebrities in videos that promote certain schemes, deep fake content generators can orchestrate financial fraud campaigns with little to no effort.
In the world of video interviews and meetings, deep fakes can also open doors to identity theft. This puts almost everyone at risk, with a bigger target on the back of higher level executives.
After learning what is deep fake, it is easy to determine that the technology can be a double-edged sword. But as long as you are aware of the risk as well as advantages of the technology, you can make an informed decision about how to consume or use media that is generated through this approach.
With that being said, it is important that deep fake meaning gets a larger exposure to everyday audiences through clear disclosures in generated media. While it may break the illusion of the technology, it will help people understand the reality of the content that they are actively consuming.