On November 16, Google unveiled its latest music generation model, Lyria, developed by Google DeepMind. This advanced AI model is designed to excel in creating high-quality music with instrumentals and vocals, as well as performing various music-related tasks while allowing users more control over style and performance.
Google also announced two AI experiments in collaboration with YouTube. The first, called Dream Track, is an experiment within YouTube Shorts. In addition to the Dream Track experiment, Google and YouTube have been working on music AI tools, allowing users to create new music or transform audio from one style or instrument to another. This includes turning vocal recordings into instrumental compositions or changing melodies into different instrument tracks.
To address copyright concerns, Google introduced a tool called SynthID, which watermarks AI-generated audio content without affecting the listening experience.
Lastly, YouTube is developing a system for rightsholders to request the removal of content that mimics an artist's voice, initially available to labels and distributors involved in YouTube's AI music experiments.
Universal Music Group's Chairman and CEO, Sir Lucian Grainge, emphasized the importance of protecting artists from unauthorized exploitation while harnessing AI's potential for creativity and commercial success.
Warner Music Group's CEO, Robert Kyncl, applauded YouTube's collaborative approach in working with artists, songwriters, labels, and publishers to experiment and find innovative solutions.
Several artists, including Charlie Puth, Sia, Papoose, John Legend, Demi Lovato, Charli XCX, and Alec Benjamin, expressed their excitement and optimism about collaborating with AI technology to explore new possibilities in music creation and shape the future of the industry responsibly.