What I Learned from EOTO Presentations – #2

From listening to my classmates EOTO presentations, one thing I found interesting to learn about was deep fakes. A deep fake is where a person's face or entire body is altered to appear as someone else. The person creating the deep fake can then take a video, and they will look like the whoever they are impersonating.  

In the presentation, I learned that the scary thing about deep fakes is how realistic they look. Deep fake videos can easily convince people they are real, are it is even hard for technology to detect if a video is fake. Despite this, there is several signs that a video is a deep fake. One sign is that the lighting is mismatch, where one section of the video seems to have different lighting quality than another. Another sign is unnatural eye movement – such as if the person in a video never blinks or blinks very often. Along with this, unnatural or awkward body movements is another sign a video is a deep fake. 

I learned that anyone is able to create a deep fake video. Originally, I thought that deep fakes were hard to make and required a certain level of expertise. However, through the EOTO presentation, I learned that anyone can easily make these videos on sites like FaceMe, FaceApp, or Wombo.  

I also discovered there are a few benefits of deep fakes. For instance, online beauty or clothing companies can use this technology to better demonstrate to customers how a product will look on them. Deep fakes can also be used by politicians to lower the cost of video campaigns, and filmmakers can use this technology when making movies. 

In contrast, there are also negative uses to this technology. One example is using deep fakes to impersonate well-known celebrities or politicians – this can potentially become dangerous if many people believe the deep fake is real. 

I find it very interesting to think about how this technology will progress and develop over the coming years. I expect the deepfake technology will only continue to get more advanced and realistic, but I also think we will develop technology to more quickly and easily tell if a video is a deep fake or not.