A recent article outlines the growing danger of deepfakes and their increased realism, due to machine learning and artificial intelligence. MIT's Center for Advanced Virtuality has created a technologically advanced storytelling project that has manipulated archived footage of President Richard Nixon's 1969 speech during the Apollo 11 lunar landing. The goal of the video is to demonstrate the convincing deepfake technology and warn the public of the burgeoning threat of media misinformation.
While the manipulation of photography has existed since the 19th century, becoming affluent during the era of motion pictures, the current state of deepfakes has become Reddit and 4Chan.
The article recently released by 1969 Apollo 11 mission. In this deepfake, President Nixon delivers the actual contingency speech written in the event that the Apollo 11 astronauts had not returned home. The goal of the video is to not only exhibit the urgent need to recognize these threats of misinformation in an era of "fake news" and copious social media, but to also showcase the technology and AI as expressions of creativity and truth. In addition to the six-minute Nixon deepfake, the moon disaster website provides educational resources on deepfakes, including a behind the scenes look at the making of the deepfake itself. You can view the trailer to the "In Event Of Moon Disaster" below.
How Dangerous Is Deepfake Technology?
For the most part, the term deepfake comes with a negative connotation. The term itself is akin to what it represents, a blending of two elements - in this case, "deep learning" and "fake." This is an understandable reputation given that, since it's emergence, deepfakes has been more often used as a cunning tool against truth and fact, and has only helped muddy the waters of disinformation online. Computer scientists, especially at MIT, would argue that this technology is not without merit, however. The advances in artificial intelligence and deep learning used within these computer technologies are an achievement in themselves, regardless of their immoral use thus far.
The MIT Center for Advanced Virtuality has recognized this breakthrough and has decided to use it for good, raising awareness through creativity and archival footage to create an alternate universe in the nation's history that very much could have been a reality. This is a creative attempt to share the technology with online s who would not seek information on deepfakes themselves or more importantly, people who have never even heard of deepfake technology.
Regardless of one's knowledge of the technology, all are still at risk of its influence if they don't learn to recognize it. In the current socially distanced world, people's work, children's schools, and how we seek and receive all sorts of daily information is via the internet and with the assistance of AI. It's an eerie sight to see a deceased President give a somber speech that never actually happened, but what's more alarming is the technology that makes it so believable. Hopefully, as more videos like this gain attention it will help to awareness of deepfakes, so that the world and its scientists will be able to better recognize it, and flag content as fake before any more humiliation or disinformation is spread.
Source: MIT News