ThisPersonDoesNotExist.com, a website created by an Uber software engineer, generates hyper-realistic portraits of people who don’t actually exist.
Uber software engineer Phillip Wang reportedly used “research released last year by chip designer Nvidia to create an endless stream of fake portraits.”
“The algorithm behind it is trained on a huge dataset of real images, then uses a type of neural network known as a generative adversarial network (or GAN) to fabricate new examples,” the Verge reported.
Every time a user refreshes the website, a new realistic face is generated.
In a Facebook post, Wang explained, “Recently a talented group of researchers at Nvidia released the current state of the art generative adversarial network, StyleGAN[…] I have decided to dig into my own pockets and raise some public awareness for this technology.”
“Faces are most salient to our cognition, so I’ve decided to put that specific pretrained model up. Their research group have also included pretrained models for cats, cars, and bedrooms in their repository that you can immediately use,” he continued. “Each time you refresh the site, the network will generate a new facial image from scratch from a 512 dimensional vector.”
Similar technology has been used to create fake pornography videos featuring the faces of celebrities on the bodies of porn stars.
Rising to prominence in 2017, “deepfake” pornography videos have featured Scarlett Johansson, Emma Watson, Gal Gadot, Maisie Williams, and Aubrey Plaza.
Deepfake technology has also raised concerns about identity theft and national security, and in 2018, BuzzFeed released a realistic, but fake, video of former President Obama saying things in a PSA that he didn’t really say.
In December, Kevin Kelly, the founding executive editor of Wired, claimed the technology marked the “end of photography as evidence.”
— Kevin Kelly (@kevin2kelly) December 14, 2018