Meta has launched a new language learning model called MusicGen, which can be used to generate AI-generated music that can be used for creating soundtracks and music tunes. Developed by the company’s Audiocraft team, MusicGen is similar to a musical version of ChatGPT. It works by entering a text description of the type of music you want and generating a 12-second track according to your instructions. The AI uses licensed music tracks and royalty-free songs to generate a wide range of musical styles.
MusicGen is currently available to the public on Meta’s Hugging Face website, but it can only produce instrumentals. It is not capable of producing vocals. However, users can upload their own song to steer the AI, enabling MusicGen to have a better sense of structure. The AI is great at generating simple melodies, but it lacks the depth and quality that human ingenuity can provide.
Musicians don’t have to worry about losing their careers, as AI-generated music has its limitations. While the AI can be useful in creating plain background audio for videos or presentations, it may not be engaging enough to create a pop hit. Meta’s model is unstable, and its website has crashed multiple times due to high user traffic. Additionally, the company may face legal challenges from record companies over illegal use of copyrighted works.
In conclusion, while AI-generated music has its place, it is not yet a viable replacement for human creativity. MusicGen can provide some exciting results, but it is not capable of producing music on the same level as a true musician. Nonetheless, users interested in trying out MusicGen should visit the Hugging Face website and experiment with the AI.