In one of our previous articles we saw how advanced artificial intelligence has become by understanding its ability to create images of fake people or individuals who do not exist. Thispersondoesnotexist.com was a perfect example to understand and guess the possible capabilities of AI in the future.
Now, Whichfaceisreal.com marks the continuation of the story started by thispersondoesnotexist.com. While the former tested the AI to generate images of fakes, the latter tests the ability of humans to identify the real from the fakes.
The generative ability of AI to come up with faces that belong to no one on the planet raises a lot of concerns including the human ability to trust digital information. Whichfaceisreal.com is aimed at upskilling humans in identifying the fakes. Fake faces generated by AI will be a huge threat to the public, especially on social media.
Jevin West and Carl Bergstrom, two academics from the University of Washington who are behind the new website were driven by these concerns. They think the rise of AI-generated fakes could be trouble and will undermine society’s trust in evidence. With Whichfaceisreal.com Jevin West and Carl Bergstrom wishes to educate the public and help them to prevent falling for fake faces on the internet.
Carl Bergstrom reportedly told a news media “When new technology like this comes along, the most dangerous period is when the technology is out there but the public isn’t aware of it,”.He added, “That’s when it can be used most effectively.”
“So what we’re trying to do is educate the public, make people aware that this technology is out there,” said West. “Just like eventually most people were made aware that you can Photoshop an image,” he added.
Is The GAN-Powered Website ThisPersonDoesNotExist.Com Creepy Or Brilliant? familiarized us with the StyleGAN technology developed by NVIDIA, the algorithm that helps machines imagine and come up with fake yet realistic faces. Whichfaceisreal.com uses the same algorithm to create fakes but for what the creators believe to be a greater purpose of helping people.
The technology is more capable, the same tech that generates images can also manipulate other media like audio or video. There is a limitation to what can be achieved now only limited by the tech available but the story will change in a few years.
Whichfaceisreal.com is essentially a game that teaches us the difference between AI generated fakes and real faces in the world. The website also explains the differences and gives tips and instructions to the users on how to spot fakes.
The website became immediately popular with about 4 million plays in just 2 weeks as reported.
An initial analysis made on the attempts by the users says almost 70% of the attempts are correct with users being able to pick real faces against fakes.
West, one of the creators said that they need to do more rigorous analysis on the attempts as it would give more insights such as which of the fakes are more realistic and difficult for the people to distinguish, images of kids or elderly people
Conclusion
Going ahead it will be harder for us humans to keep up with the fakes generated by AI. Although attempts like whichfaceisreal.com may educate and spread the word about what our technology is really capable of it can do very little in stopping the tech from being misused. Next time when you get a friend request from Facebook or a said match on tinder you will have to think twice or thrice before even assuming that he or she is real. As of now the only way to check if a face is real or not is by reverse searching on Google. If you find multiple images of the same face, at least you can be sure that the image is of a real person though you can’t be sure if that’s the person on the other end.
