Google set out to completely transform the future of music creation. While it may seem ambitious, it’s a plan that I’m more than convinced of when I see (and hear) evidence of how AI could create art.

The company just announced Lyria via Google DeepMind. The most advanced AI music creation model to date. For now, you’ll be introducing this with two main tools: Dream Track and the Music AI Tools lab.

Lyria is Google’s new model for producing music through artificial intelligence. It goes far beyond what we can currently test in Google’s experimental lab. MusicLM, presented a few months ago, was a musical “ChatGPT”. quicksong created.

The first taste of Lyria’s potential will come via YouTube Shorts via the Dream Track tool. In this way, we will need to enter the prompt as in MusicLM. The main difference is that in this case they establish alliances with artists Like Alec Benjamin, Charlie Puth, Charli XCX, Demi Lovato, John Legend, Sia, T-Pain, Troye Sivan and Papoose. They will be the ones “singing” the songs with the command we enter.

Beyond this interesting and powerful tool, the most amazing thing is Music AI Tools. A simply brutal music creation interface, as we see in Google examples. These tools have the ability to turn a humming sound into an orchestral melody or any instrument we want.

We can follow the process in reverse by turning the chords of a MIDI keyboard into a vocal chorus, or turning a few seconds of beatboxing into a drum loop for an instrumental. Essentially, creating complex musical compositions with any tool we have at our disposal, be it vocal or instrumental.

Currently the tools are not active in AI Lab and Dream Track is not available in Spain via Shorts, but we will start to see them implemented in the coming weeks.

Image | Microsoft Bing Creator and Photoshop

in Xataka | The Beatles are releasing their latest songs thanks to artificial intelligence. A premiere that doesn’t shy away from ethical debates about using the voice of a dead person