ThatsWhyHow logoThatsWhyHow
AI

How can a computer copy someone's voice?

Short answer: An AI model learns pitch, tone, accent and rhythm, then says anything in that voice. Here's how it works, and the FTC's advice for spotting a fake emergency call. So voice cloning learns the sound of a voice and lets a computer say anything with it.

Full transcript

Hook

Why and how can a computer copy someone's voice? AI voice clones are getting so good that scammers use them to fake phone calls, even calls to 911.

Learn

First, an AI model studies recordings of a person speaking. It learns what makes that voice unique, like pitch, tone, accent and rhythm.

Sample

Older systems needed hours of audio. Newer ones can copy a voice from a short clip, like a video posted online.

Speak

Then you type any sentence. The model turns the text into speech, building the sound one tiny slice at a time, in the cloned voice.

Good

The same tech has real uses. It can help people who are losing their voice to illness keep sounding like themselves, and it narrates audiobooks and videos.

Scam

Scammers use it for fake emergencies. The call sounds exactly like a family member in trouble, begging for money right now.

Protect

The FTC says don't trust the voice. Hang up and call the person back on a number you know is theirs. And be wary of anyone demanding wire transfers, crypto or gift cards.

Close

So voice cloning learns the sound of a voice and lets a computer say anything with it. ThatsWhyHow.

Sources

Spotted an error? Leave a comment on the video. Corrections are pinned and noted here.