On Wednesday, Google unveiled an exciting new feature for its Gemini app that allows users to generate music. This innovative capability is powered by DeepMind's Lyria 3 music generation model, which is currently in its beta phase.
Users can create their own songs by simply describing the type of track they wish to generate. For example, one could request a "comical R&B slow jam about a sock finding its match," and the app will produce a 30-second music track complete with lyrics and cover art designed by Nano Banana.
Furthermore, the app enables users to upload a photo or video, prompting the AI-driven tool to compose a song that reflects the mood of the media file.
Google highlighted that Lyria 3 represents a significant improvement over its predecessors, delivering more intricate and realistic music tracks. Users have the flexibility to adjust various elements such as style, vocals, and tempo to tailor their music experience.
In addition to integrating Lyria 3 into the Gemini app, Google is also expanding access to YouTube creators via the Dream Track feature. This tool assists content creators in producing AI-generated tracks, which was previously limited to U.S. creators but is now available globally.
While users cannot replicate an artist's work directly, they can mention an artist's name in their prompts, allowing Gemini to generate music in a similar style or mood. Google clarified that the aim of the music generation feature is to foster original creativity rather than imitate existing musicians. The platform includes safeguards to ensure outputs do not infringe on existing content.
All tracks produced using the Lyria 3 model will feature a SynthID watermark to signify their AI-generated nature. Additionally, Google is enhancing capabilities within Gemini to identify AI-generated music, allowing users to upload tracks and inquire whether they are AI-generated.
This music generation feature will be rolled out to all Gemini users aged 18 and over, supporting multiple languages including English, German, Spanish, French, Hindi, Japanese, Korean, and Portuguese.
The emergence of AI-generated music has sparked various reactions among artists and audiences. While platforms like YouTube and Spotify are embracing AI technology and forming partnerships with music labels to monetize these tracks, there are ongoing legal challenges from the music industry regarding copyright issues related to training materials used in AI development.
