With several lawsuits already in motion, it remains to be seen how the law will address AI-generated music and how that may affect further innovation.
Meta has launched a new AI music generator MusicGen that can create original songs from text inputs and existing melody prompts. Felix Kreuk, a member of the Meta Fundamental AI Research (FAIR) team, announced this via his Twitter page.
Recall that in March, Meta CEO Mark Zuckerberg announced the company’s increased efforts in AI development for the Metaverse. He said, “Our single largest investment is in advancing AI and building it into every one of our products.”
MusicGen is the latest example of how Meta is leveraging its AI expertise to create innovative and immersive user experiences. Previously, the social media giant launched DALL-E, an image generator that can draw anything from text inputs. It also unveiled ReBeL, a reinforcement learning algorithm that can master any two-player game. Meta hopes to use AI to empower people and connect them through meaningful interactions.
How the AI Music Generator Works
According to Meta, MusicGen has been trained on 20,000 hours of music across different genres. The AI can mix various elements of these music styles to generate high-quality audio files based on user specifications.
Unlike Google’s MusicLM, MusicGen is available for free on GitHub, where users can download the source code and run it on their own devices. According to Felix Kreuk, the team has made it available publicly “for open research, reproducibility, and for the music community”. MusicGen also has a web interface where users can try it out online and share their creations with others.
Ethical Concerns
MusicGen could become a valuable tool for music enthusiasts, musicians, composers, and producers who want to generate ideas, experiment with different genres, and enhance their workflow.
However, some critics have raised concerns about the ethical implications of MusicGen and other AI music generators. They argue that such tools could pose a threat to the originality and authenticity of music, as well as the rights and income of musicians. Nonetheless, there is still a lack of clarity on how AI-generated music may violate the copyright of artists, labels, and other rights holders.
With several lawsuits already in motion, it remains to be seen how the law will address AI-generated music and how that may affect further innovation.
An experienced writer with practical experience in the fintech industry. When not writing, he spends his time reading, researching or teaching.
