Deep Fakes and RealTalk: AI Simulating Humans

MarriedWarmheartedArrowcrab-max-1mb.gif

If you thought click bait and fake news is bad now, wait to you see what’s coming. The future is dangerous. The rapid advancements in technology continue to bring things from our imaginations and sci-fi movies to life and not without scary repercussions. Oh you thought 3D printing working firearms was from ‘Lost in Space’… Well, you’re wrong.

But that’s not the point. How can fake news get any worse? Well it’s great you asked because with Deep Fakes and RealTalk, it’s gonna be a shit show.

So WTF is this, right? Simply, RealTalk is a Deep Synthesis Model made by Dessa, that replicates human speech.

The replica of Rogan’s voice the team created was produced using a text-to-speech deep learning system they developed called RealTalk, which generates life-like speech using only text inputs.

And it only gets worse.

This is a Deep Fake

a technique for human image synthesis based on artificial intelligence. It is used to combine and superimpose existing images and videos onto source images or videos using a machine learning technique called a "generative adversarial network" (GAN).[2] The phrase "deepfake" was coined in 2017.

So what you’ve just seen is AI simulate a human voice AND human face.

1J5sfaZ.gif

Now this isn't particularly new nor are these the only technologies capable of creating these “effects.” In the Samsung AI lab, they've successfully animated a photo, using only a few shots of a person. Here's the paper on it.

Now, think about it. This is a major issue because not only will there potentially be more fake news but the how will we be bale to know the difference? It’s scary. Now, there are a lot of things this kind of technology can help, like:

  • Talking to a voice assistant in a way that feels as natural as talking to a friend

  • Customized voice applications — for instance, a workout app that contains a personalized pre-workout pep talk from Arnold Schwarzenegger

  • Improved accessibility options for people that communicate through text-to-speech devices, for example, people with Lou Gehrig’s disease

  • Automating voice dubbing for any media and in any language

But in the wrong hands, spam calls, impersonations, fake images and so much more fake news. The good news is that to complete something like this, technical know how, ingenuity, computing power and data are required to these things work and not just anyone can just go and do that.

The future is scary and know that things are just going to get crazier.