Can deepfakes be detected?
Hello, everyone.
I welcome you to my blog. With each passing day there is a technological development that is being worked on or is being introduced into our world; we have come this far because no day passes without new developments being introduced or worked on. Once upon a time we depended completely on animals for transport; it was slow and not always reliable, but with the help of technological advancements and developments, we got cars, trains, and other means of transportation. Though back then the cars were not that fast and not that comfortable, it was better and more reliable than animals.
Today we have super fast cars, lorries, trains, aeroplanes, and ships; they are fast, reliable, and most of all, they are comfortable. Those days you cannot sit for hours in these cars and trains without getting exhausted, but over the years, with the help of the continuous evolution of technology, we have been able to come up with transport systems where we can be seated for hours without even feeling our backs ache or feeling tired or exhausted. Every day, technology is working towards making life more comfortable for man, and that is how we came this far and have all that we have today.
The continuous evolution of technology brought about artificial intelligence, and artificial intelligence through continuous evolution brought about deepfakes. Today, you as an individual can see yourself in a video or picture you don't even remember making or taking, and it is all thanks to artificial intelligence. On Twitter, a girl posted different pictures of her and her father, two distinct pictures, and asked that AI put both pictures together, and the result is something that anyone will see and find it difficult to believe that both pictures were not together from the very beginning.
The other day I also saw a video on TikTok. Pep Guardiola, the manager of Manchester City, was saying a lot of bad things about Arsenal, and I was left wondering why he would say such a thing and not get sanctioned by FIFA. The video looked so real that even I was convinced until I went through the comment section and I saw comments where other users were saying the video was AI-generated.
After realizing it was AI-generated, that is when I started noticing the inconsistencies in the video, and how his mouth was moving showed that what was said in the video was not actually said by him. Deepfakes might be difficult and challenging to detect, but they can actually be detected.
With the current technology level, we do not have much at our disposal that we can use to detect deepfakes, but through a close look at the picture or video, we can tell if it was AI-generated or not. This might be difficult to tell, but with a close look you can detect some inconsistencies in the video or picture; you can notice the movement or the background and from there deduce that they are not real and are AI-generated.
Another way to detect deepfakes is with the use of AI tools. There are a few AI tools that have been designed to help detect deepfakes; this makes it a bit easier and faster to deduce and conclude that a picture or video was generated using AI.
