Deepfakes — The Democratization of Forgery
At the end of last year, so-called “deepfakes” made headlines all around the world. A Reddit user had posted pornographic videos in which the actresses’ faces had been replaced by the faces of celebrities like Emma Watson and Taylor Swift. How did he do it? You guessed it: Artificial intelligence!
Needless to say, deepfakes quickly gained a substantial amount of both fans and critics. Although a fantasy come true for some, in the wider public a debate about the ethical implications of such videos emerged, soon reaching the consensus that they were in fact bad.
PornHub, Twitter, Gfycat and many others were quick to announce that they would not tolerate such content on their platforms, Reddit banned a community of about 50,000 teenagers who had mysteriously found a new passion for artificial intelligence, the panic was over and the world turned its attention back to other important issues.
But the story of deepfakes doesn’t end here. In fact it is only just beginning. With the technology still being in its infancy, porn was simply the first real-world application popular enough to gain widespread attention. But as their quality rapidly improves and their hardware requirements shrink, deepfake AIs are on track to conquer many other domains in the near future.
What is currently taking experts several hours, if not days to produce, using professional-grade hardware and software, will soon take an AI mere seconds. The CGI of tomorrow will be produced by a 14-year old tapping a few buttons on their smartphone and achieving results that are indistinguishable from reality. For the first time in history, faking things will be laughably easy.
What we are about to experience is the democratization of forgery. And nobody is prepared for it!
Photos, videos, audio recordings provide us with evidence that a given thing happened, they are proof of our experiences, of what we saw and heard. The phrase “pics or it didn’t happen”, although usually used rather tongue-in-cheek, captures this idea perfectly: Images, whether still or moving, are the highest authorities of truth. They are the definition of reality. They are so essential to our information ecosystem that it is hard to imagine a world where we could not trust in their veracity anymore.
But that is exactly the future we are approaching. With everyone being able to create perfect forgeries, how can anyone know what is real and what is not? How does a justice system work in which images cannot be used as evidence? Who can a journalist rely on in a world where trust has become a rare commodity?
Over the course of photographic history humanity has trained itself to trust images as depicting reality. In the future, however, our complete loss of trust might actually allow us to be more critical thinkers, no longer blindly trusting a medium that has been manipulative from its very beginning.
In her book On Photography, Susan Sontag writes: “Humankind lingers unregenerately in Plato’s cave, still reveling, its age-old habit, in mere images of the truth.”
Who knows, deepfakes might just be what frees us from Plato’s cave.
