Deepfakes AI Meets Social Science
摘要
Humanity knows and understands that “there is nothing more important than truth”. But effect of truth is much lesser when there is a spread of fake news, and when that fake news looks to be authentic, reads authentic, sounds authentic, and comes from authentic sources. This is referred to as deepfake. Although the history of the word” deepfake” is not that deep, (firstly coined in 2017 by reddit user), the spread of visual disinformation is historically old and goes back to the era of 1940’s in Soviet Union propaganda. With the advancements in the field of deepfakes and the technology of its creation and detection, it is becoming harder to fight against these sources of misinformation. This study aims to look at the social impacts of deepfakes, its types, and the technology surrounding deepfake detection and generation.