Youtube is testing a new AI tool – create custom music for your Shorts

Have you ever wanted to upload a YouTube Short video, but always struggled with the right music? Well, it seems that AI is about the solve this problem as well. Reports have started coming out of YouTube HQ that there are active tests ongoing on a new AI tool, which will enable creators to make their own 30-second music to add to their Shorts. They are also seemingly working with a few mainstream artists, who have agreed to lend their voice to this tool. So, how will this work?

Once released, the creator will be able to use the AI tool to create custom audio for their shorts content. This will be done via a prompt, a vocal sample, or a sampling of the voice of the pre-installed artists. Among the singers who have agreed to lend their voice are popular names like John Legend, T-Pain, and Sia. The project has been named Dream Track and while it was announced a while ago, it seems that the negotiations with the record labels of the participating artists have concluded now and that has made it possible for YouTube to go ahead with final testing.

 

 

YouTube will be releasing Dream Track to a select group of creators in America, followed by a gradual release to other users. There is no news yet on whether we will have a stand alone package or a desktop version of the Dream Track application.

If you are interested in seeing this tool in action, there is a video on YouTube, where they have demonstrated a track being created using a prompt and the voice of T-Pain. It is admirable that the tool is able to mimic the style of T-Pain in terms of the lyrics, but the audio output itself is a work in progress. The final build of the Dream Track app will be much better as they will spend more time training a dedicated model for each participating artist.

 

 

Another AI app that is in the works at YouTube is called Music AI Tools. This one relies on the user supplied vocal sample, that it can morph into audio of different styles. For example, you can upload a vocal sample of yourself humming a tune, and the Music AI Tool will be able to convert that into a guitar strumming.

There are a few images online showcasing what the interface of this tool will look like. There is not a lot of noise about this at the moment as YouTube has confirmed that it is still in its nascent stages and there is still some work to be done.

 

YouTube

 

It will be interesting to see how fast YouTube can get these tools into the market, but one thing is for sure – Content creation will never be the same again.