Send in the Clones

Uncanny Valley | WIRED

WIRED

Technology

4.1 • 572 Ratings

🗓️ 13 April 2023

⏱️ 38 minutes

🔗️ Recording | iTunes | RSS

🧾️ Download transcript

Summary

Artificial intelligence continues to seep into every aspect of our lives: search results, chatbots, images on social media, viral videos, documentaries about dead celebrities. Of course, it’s also seeping into our ears through our podcast clients.

A new class of emerging AI-powered services can take audio clips from voice recordings and build models off them. Anything you type into a computer can be spit out as an impression of that person’s voice. Proponents of AI voice cloning see these tools as a way to make life a little easier for content creators. The robo-voices can be used to fix mistakes, read ads, or perform other mundane duties. Critics warn that the same tools can be weaponized to steal identities, scam people, and make it sound like someone has said horrible things they never did.

This week, we ask our producer Boone Ashworth, who is also a staff writer for WIRED, to sit down in front of the microphone and bring his AI voice clone experiments with him.

Show Notes:

Read Boone’s story about AI voice clones. Read all of our recent coverage of artificial intelligence.

Recommendations:

Boone recommends the Arte Concert Passengers playlist on YouTube. Mike recommends The New York Times Presents: The Legacy of J Dilla, a documentary on FX and Hulu. Lauren recommends starting Succession—From the beginning! Very important!—if you haven’t. It’s on HBO.

Boone Ashworth can be found on Twitter @BooneAshworth. Lauren Goode is @LaurenGoode. Michael Calore is @snackfight. Bling the main hotline at @GadgetLab. The show is produced by Boone Ashworth. Our theme music is by Solar Keys.

Learn about your ad choices: dovetail.prx.org/ad-choices

Transcript

Click on a timestamp to play from that location

0:00.0	Mike.
0:01.0	Lauren.
0:02.1	Do you think this is my real voice?
0:05.4	Yes.
0:06.3	I'm pretty certain it's your real voice.
0:09.2	Why is that?
0:09.9	Vocal fry.
0:10.5	It was the vocal fry that gave it away, huh?
0:11.9	Well, no, it's the fact that I'm watching your lips move.
0:13.7	But if I shut my eyes and listen to you speak, I think I would still think it was your voice speaking.
0:19.1	Wow.
0:20.1	What you just said was really poetic.
0:23.2	It sounds like something that Samantha, the AI, in her, the movie her, would say at the end.
0:29.7	Say that again, but slowly.
0:31.0	And in my Scarlet Johansson voice?
0:33.2	It's kind of like, but I shut my eyes and listen, I would still think it's actually you talking.
0:38.0	It's like at the end of the film when she talks about the pages of the book getting further and further apart.
0:41.9	Oh, sure. Yeah. Do you think in the future we could maybe simulate Scarlett Johansson's voice to host this podcast and then you and I are off hosting duties?
0:51.7	I mean, I enjoy doing this show every week, but that would be a future that I don't think is
0:57.0	possible.
0:57.8	I mean, it is possible.
0:58.8	This show will just be two robots talking to each other in the future.
	...

Please login to see the full transcript.

Previous episode | Next episode

Disclaimer: The podcast and artwork embedded on this page are from WIRED, and are the property of its owner and not affiliated with or endorsed by Tapesearch.

Generated transcripts are the property of WIRED and are distributed freely under the Fair Use doctrine. Transcripts generated by Tapesearch are not guaranteed to be accurate.