YouTube is pushing the boundaries of music creation with artificial intelligence. Building on its “Dream Track” initiative launched last year, the platform is now testing a new feature that allows select creators to remix songs using AI.
This innovative tool, called “Restyle a Track,” lets creators reimagine existing songs by providing textual descriptions of their desired style. Want to transform a pop song into a high-energy dance track? Or perhaps infuse a ballad with a touch of rock and roll? Simply input your vision into the prompt, and YouTube’s AI will generate a 30-second remix snippet ready for use in Shorts.
YouTube emphasizes that these AI-generated remixes will maintain the essence of the original song’s vocals and lyrics while adapting the overall style. Proper credit will be attributed to the original artists through the Short itself and the Shorts audio page. Additionally, clear labeling will indicate that the track has been modified using AI.
This new feature expands upon YouTube’s existing Dream Track toolset, which enables creators to generate AI music from scratch using prompts or even by humming a tune. These tools are powered by DeepMind’s Lyria music generation model.
Recognizing the importance of fair compensation in the age of AI music, YouTube has proactively sought agreements with artists and rights holders. The company has partnered with Universal Music Group (UMG) to develop a framework for paying rights holders when their music is used in AI-driven features.
YouTube is not alone in exploring AI-powered music remixing. Other platforms, such as the upcoming app Hook, are also developing tools to empower creators with new ways to reimagine and personalize music.
This latest development from YouTube highlights the growing potential of AI to democratize music creation and provide creators with innovative tools to express their artistic vision. As AI technology continues to evolve, we can expect even more groundbreaking features to emerge, transforming the way music is produced, shared, and experienced.
