It Only Takes a Few Dollars and 8 Minutes to Create a Deepfake
If you’ve ever heard of deepfakes, you know how frightening and awe-inspiring they are. Recently, a Wharton Professor created a deepfake of himself to show how easy it was to manipulate video content. It cost him only a few dollars and eight minutes to create a deepfake video that was an exact replica of himself. This begs the question – if it’s this easy, then how can we trust what we see and hear on the internet? Let’s take a deeper dive into deepfakes and their implications.
Deepfakes use a neural network to mimic the speech, facial expressions, and body movements of a real person. This process involves feeding a computer with hundreds or even thousands of images of the individual, and then coding it to transform and blend one video’s characteristics to another. The resulting deepfake is almost indistinguishable from the original video.
Creating a deepfake used to be a time-consuming, expensive process that required a high level of technical expertise. However, with the advent of machine learning algorithms, it’s now possible to make a deepfake in just a few minutes using off-the-shelf software and a few dollars.
Deepfakes pose a significant threat to ordinary people and public figures. They can be used to defame someone, ruin their reputation, or manipulate them into doing something they wouldn’t otherwise do. The obvious victims in these scenarios are politicians, celebrities, and influencers with large online followings. However, deepfakes can also affect everyday people. For example, a deepfake could be made about an individual causing reputation damage or even harassment.
The risks of deepfakes have sparked a global conversation about how they can be controlled. Lawmakers in some countries have moved to put in place legislation to regulate the creation and distribution of deepfakes, while tech companies are trialing AI models that detect deepfakes. However, regulating deepfakes can be challenging. Once a deepfake is created, it can be shared worldwide in a matter of seconds, making it hard to trace the origin.
In conclusion, deepfakes are a new technological challenge that poses risks to individuals, organizations, and governments alike. Although research is ongoing in the control of them, it is troubling to see how easy they can be created and manipulated in comparison to the technics used to create the early ones on Dali’s movies. In this era of technological innovation, we need to be vigilant and take the necessary steps to ensure that deepfakes aren’t used to deceive, harm or discriminate against anyone. It’s essential to practice good media literacy and to look closely at the content we consume to ensure that it’s legitimate.