YouTube tests AI tools to clone pop singers’ voices

Credit: youtube

YouTube has announced that it is testing new generative AI tools that will allow creators to create their own music and songs from prompts.

The first of the two tools is the Dream Track, which will allow users to use the voices of nine famous pop singers to make their own songs.

The nine artists who participated, as part of a collaboration with YouTube, in the project are Alec Benjamin, Charlie Puth, Charli XCX, Demi Lovato, John Legend, Papoose, Sia, T-Pain and Troye Sivan. 

The new tool will allow users to create songs that are based on these artists’ styles and voices.

These songs will be available to use on the platform’s short-form videos, YouTube Shorts.

The platform also launched two videos to show an example of how the tool would work using Charlie Puth and T-Pain’s voices.

At the moment, access to this tool has only been given to a small group of US content creators.

While the tool uses the voices of famous artists, YouTube also emphasises the responsibility that comes with the use of AI.

In a statement, it said, “Despite the tremendous opportunity AI presents, we also recognise it’s a quickly evolving space that presents complex challenges.”

The new tool seems to be an effort to maintain the platform’s relationship with music labels while exploring AI-generated music,

“One of YouTube’s greatest strengths is our strong relationships with music industry partners.

“We’re committed to collaborating with them as we enter this new era, critically exploring together new opportunities and developing sensible and sustainable controls, monetisation and attribution frameworks,” YouTube added.

YouTube also revealed its other AI offering, Music AI Tools which will allow users to create songs from humming.

The prompts will then allow the user to convert the humming to any form of instrument.

It launched a video to showcase how the range of tools would work.

Both tools have been created in collaboration with Google’s DeepMind.