Meta (formerly known as Facebook) has unveiled its latest development in the art of music, MusicGen. This is a large language model developed by the Audiocraft team that allows you to create short tracks using artificial intelligence (AI).
- MusicGen is capable of creating tracks up to 12 seconds long using text prompts provided by the user.
- Users can “guide” MusicGen by providing their own songs so that the model better understands the query structure.
- This tool can be useful for creating background audio for videos, presentations, and other projects.
- Unlike Google’s MusicLM AI music generator, the MusicGen model cannot perform vocals.
- MusicGen does a good job of creating simple, short tunes, but it’s not designed for more complex compositions.
Use of licensed tracks and copyright issues:
- MusicGen was trained using licensed music tracks as well as free songs from Shutterstock and Pond5.
- As the use of copyrighted works becomes more problematic, Meta and MusicGen may become the subject of scrutiny and possible lawsuits from musicians and copyright advocates.
MusicGen availability and trial:
- MusicGen is currently available to the public on the Hugging Face website.
- Experimenting with MusicGen can help users create background audio for their projects.
With the advent of MusicGen Meta provides the user with a convenient tool for creating short music tracks, concludes NIX Solutions. However, problems with the use of licensed tracks and copyright protection could potentially arise in the future. It is important to pay attention to the legality of the use of music content and ensure that copyrights are respected when creating music using artificial intelligence.