Take a good look at the collage of people here. Now click on the image and take a really closer look and tell me the one thing all these fine folks have in common with one another.
Give up? Okay, I’ll tell you then: none of them are real. These peoples’ photos have been computer generated using artificial intelligence and technology developed by graphics card producer, Nvidia. Don’t believe me? Head over to the website ThisPersonDoesNotExist.com and give it a shot yourself. Once the image is loaded and you’ve had a chance to check it out, refresh your browser and another “person” will be generated from scratch.
Now, I know you’re probably thinking, “Okay, well, what’s the big deal? I’ve seen lots of computer generated ‘people’ before…” Think Audrey Hepburn in the recent chocolate commercial or Carrie Fisher or Peter Cushing in the Star Wars film. But these incarnations are very different in that there are a lot of incredibly talented people (artists, animators, programmers, etc) behind the scenes who spend tons of time and money to create this digital likeness. What’s different about the photos of the people generated on ThisPersonDoesNotExist.com here is that a computer is creating these images from scratch, on the fly each time someone refreshes the browser window. No artists, etc needed here just some powerful computing and cutting-edge clever programming utilizing AI and GAN or Generative Adversarial Network, which uses two neural networks.
Well, that’s certainly cool, right? I mean many folks might already be thinking how much more realistic this means videogames will be within just a couple of generations (videogame platform generations like Xbox, PlayStation, and Switch, not human generations). Sure, and that’ll be really cool, too!
Others might be thinking about the wider, more dramatic changes that technology like this might usher in. Think Hollywood. Specifically, think Hollywood actors. While the idea of virtual or digital actors isn’t new, this incredibly realistic technology is. The idea of virtual entertainers isn’t anything new either. Just take a look at a recent (2018) concert in Los Angeles in which a virtual Japanese “idol”, Hatsune Miku, performs for a live audience. She’s basically a hologram and her voice is synthetic. Yet, look and listen to the crowd – they know the songs, the lyrics, and they respond to Miku. Now imagine how much potential money Hollywood (or anyplace else that churns out entertainment) could make from utilizing this technology in the future. Actors who never age or die unless the role calls for it – and they’re certainly able to be resurrected at any time. It will provide an endless supply of new talent from which to choose.
Now, let’s go a step further…what else could something like this be used for? Like any technology, it could be used for good or bad and I think it’s prudent to assume this type of technology will be used for many “bad” things just as much or more than it will be used for good or innocent applications. How about something like fake news? It’s already becoming very difficult to determine fact from manufactured fiction sometimes and technology like this will make it even more difficult and challenging. Think about the implications for a minute. What if you, the viewer, are truly 100% unable to determine if something you see is real or not? Now think about the wider implications? What if people in general all over the world are truly unable to determine truth from fiction in the future…
Cool technology? Yeah, it is! Does it scare you? Even a bit? If it doesn’t, perhaps it should.