A few weeks ago, I blogged the AI video of Joe Rogan that flawlessly recreated his voice:
For a Joe Rogan fan like myself, the video above is hilarious, but also a little bit scary when you consider the negative ramifications.
SOURCE-Samsung researchers have been particularly adept at this, with the latest example being the partnership between Samsung’s AI research center in the UK working with Imperial College London to create an AI that animates and syncs up an audio clip with facial movements derived from little more than a photograph of someone.
One of the videos featured Rasputin, who does not get enough credit for being one of the craziest motherfuckers of all time:
Is it perfect? No, but it’s pretty damn close. They did all that from a simple photo of Rasputin. Imagine what they can do with a clear photo of YOU.
Samsung also made a video of Albert Einstein, but with audio of something he actually said:
All of that is done via the creation of what’s called a generative adversarial network, or GAN for short. “The videos generated using this model do not only produce lip movements that are synchronized with the audio but also exhibit characteristic facial expressions such as blinks, brow raises, etc.,” the researchers write in an academic paper about their work, available here. “Our improved model works on ‘in-the-wild’ unseen faces and is capable of capturing the emotion of the speaker and reflecting it in the facial expression.
Technology is good. I know that’s not really going out on a limb or anything, but technology is in fact good. However, it is scary. What happens when this takes a massive step forward and it’s an app? Are we not going to be able to believe anything? Imagine the stuff Deadspin would make on us. Also, who is going to regulate this bad boy? The government? I hope not! Or will the robots just be controlling the world by then?
The future is scary. However, if it wasn’t, I guess that would mean it would be boring. Nothing worse than boring.