Answer: A tool that lets you change what someone says just by typing it.
Good thing someone just invented that.
A research team composed of researchers from Adobe Research, Princeton University, Stanford University and the Max Planck Institute for Informatics recently released a paper presenting a new tool for working on deepfake videos. It allows a user to “edit talking-head video based on its transcript to produce a realistic output video in which the dialogue of the speaker has been modified.”
The tool currently only works on “talking-head videos” and the results can vary depending on how many words are edited. In examples shared on YouTube, the team posted a video of someone saying, “I love the smell of napalm in the morning,” followed by a deepfake version that replaced the word “napalm” with “French toast.”
According to Mashable, the researchers have not made their tool available to users and have acknowledged the ethical concerns behind doing so. While such a tool would have many useful applications in areas like movie production, it also has the potential for misuse as it could be used to spread misinformation.
Looking for the latest gov tech news as it happens? Subscribe to GT newsletters.