DEEPFAKESCultivating Media Literacy in the Age of Deepfakes
People turn to digital media for news at high rates, but at the same time algorithms for manipulating media continue to grow more powerful. Online course from the MIT Center for Advanced Virtuality seeks to empower students and educators to critically engage with media.
While people turn to digital media for news at high rates, algorithms for manipulating media continue to grow more powerful. In a Pew Research Center survey (August/September 2020), 53 percent of adults in the United States say they get news from social media “often” or “sometimes.” People have long been aware of phenomena such as “doctored” photos and misinformation at large, but machine learning is enabling the proliferation of “deepfakes,” videos or images of fake events with increasing sophistication.
Now, the MIT Center for Advanced Virtuality (MIT Virtuality for short) has created a course that addresses misinformation both in terms of specific contemporary technological phenomena and a broader media perspective.
“We are currently experiencing an information crisis,” says Joshua Glick, education producer for this MIT Virtuality project and an assistant professor of media studies at Hendrix College. “A combination of political, technological, and economic forces has propelled the spread of misinformation and disinformation throughout our media environment — and the crisis has only been amplified by the pandemic.”
The MIT Center for Advanced Virtuality, part of MIT Open Learning and directed by Computer Science and Artificial Intelligence Laboratory professor D. Fox Harrell, has created a free online course, Media Literacy in the Age of Deepfakes, with the goal of giving educators and independent learners the resources and critical skills to understand the threat of misinformation. In addition to teaching participants how to decipher fact-based assertions from lies and credible sources from hoaxes, the course aims to place deepfakes within a larger history of media manipulation and to show how activists, artists, technologists, and filmmakers are using AI-enabled media for a wide range of civic projects.
The course is implemented as a dynamic, multi-layered website including video and case study materials, as well as offering much more context and information through different self-paced learning modules. An illustrative example used in the course is “In Event of Moon Disaster,” an Emmy-winning MIT Virtuality production co-directed by Francesca Panetta and Halsey Burgund. The project showcases a deepfake of President Nixon delivering the real contingency speech written in 1969 for a scenario in which the Apollo 11 crew was unable to return from the moon, and features numerous learning and analysis resources.