Scientists at Stanford are doing their part to create what will be an inevitable dystopian nightmare.
The staff at the Max Planck Institute for Informatics, Princeton University and Adobe Research have developed software that allows you to now edit and change what people are saying in videos, allowing anyone to edit anybody into saying anything, according to Observer.
The software uses machine learning and 3-D models of the target’s face to generate new footage which allows the user to change, edit and remove words that are coming out of a person’s mouth on video, simply by typing in new text. Not only that, the changes appear to have a seamless audio/visual flow without cuts.
Here’s a video of the frightening software at work.
We’re sure there will be absolutely no blowback at all to this. After all, just last week, there was public outrage with somebody jokingly edited a video of Nancy Pelosi to make her seem drunk. What would happen if somebody edited a video of her speaking to have her swear wildly, or say racist things?
This deepfake software is already being described as “the equivalent of Christmas coming early for a Russian troll farm”, now that the 2020 election is underway. We’re sure it’ll eventually also be a topic du jour on MSNBC and CNN if Trump wins again in 2020.
And we have to ask: how long before the software is incorporated into Adobe‘s retail video editing software? After all, the software company already forces users to read a massive disclaimer that states:
We also believe that it is essential to obtain permission from the performers for any alteration before sharing a resulting video with a broad audience.
We acknowledge that bad actors might use such technologies to falsify personal statements and slander prominent individuals. We are concerned about such deception and misuse.
Are they covering themselves legally for this “technology” to go mainstream?
Meanwhile, joke deepfakes continue to pop up, like this fake video of Mark Zuckerberg sitting at a desk giving a nefarious sounding speech about Facebook‘s power.
Joe Rogan was also victim to a deepfake by the AI company Dessa recently, who released audio making it sound like he is discussing chimpanzee hockey.
Don’t worry though, we’re sure this won’t fall into the wrong hands.
June 13 (Reuters) – Twitter Inc on Thursday said it removed thousands of accounts linked to coordinated, state-backed activities it believes were from the Iranian government and archived them to its public database launched last year.
In the latest purge of information, the company said it believes 4,779 accounts were associated or backed by Iran.
The micro-blogging site also said it had removed and archived four accounts affiliated with the St. Petersburg-based Internet Research Agency, a Russian “troll farm” that has been indicted by U.S. Special Counsel Robert Mueller for attempts to interfere with the 2016 presidential election.