Tanneness got the code it's going to launch.

It's a unit system.

I know this.

It's all the files of the whole park.

It tells her everything.

Sir, he's uploading the virus. Eagle one. The package is being delivered.

Today we're going to be talking about deepfakes.

Audio deepfakes.

In 2017, Motherboard Reporter Semple first wrote about deepfakes,

which used much in learning to put the faces of celebrities on the bodies of porn actresses.

New similar technology can let us fake someone's voice.

So, anyways, we'll be talking about how that's possible.

I'm Ben Maku, and this is cyber.

No, really, I'm Ben Maku, for real.

Press me.

Okay, that was obviously not Ben Maku.

But for this episode, we're exploring the nature of deep fake audio by, of course, making a deep fake of our beloved host.

I am, for real, motherboard editor-in-chief Jason Kebler, and this is Cyber.

Ever since we started reporting on deep fakes in late 2017,

we've wanted to make a deep fake to explore how they could be used.

The technology, which relies on machine learning and generative adversarial networks,

has rapidly improved in the last few years. It's gotten tons of headlines.

...