OpenAI is developing its own artificial intelligence (AI) music generator capable of creating audio from text and audio prompts. The company is recruiting students from the prestigious Juilliard School to help train the system.
Juilliard students were required to annotate musical scores, according to a information report published on October 24. This annotated data will teach the AI to understand musical structure, pitch, and rhythm.
The tool could allow users to add guitar accompaniment to existing vocal tracks or generate original soundtracks for videos. A source familiar with the project told The Information that advertising agencies could use it to brainstorm lyrics, create jingles from musical samples or combine video styles.
OpenAI has not confirmed when it plans to release the tool. The company has also not revealed whether it will work as a standalone app or integrate with existing products like ChatGPT or video generation tool Sora. OpenAI declined to comment on the reports.
OpenAI will soon compete directly with other generative music players. Google launched the second generation of its Lyria model in May, which customers can access through Google Cloud for tasks such as creating advertising soundtracks. Startup Suno, which sells subscriptions to an artificially intelligent music generator, generates about $150 million in annual recurring revenue, almost four times more than a year ago.
This isn’t OpenAI’s first foray into music generation. The company developed two previous models, MuseNet in 2019 and Jukebox in 2020, but neither of them are currently available through ChatGPT. More recently, OpenAI has focused its audio work on text-to-speech and speech-to-text capabilities.
Copyright concerns loom over AI-generated music. The Recording Industry Association of America has sued competitors Suno and Udioalleging that they trained their models with copyrighted songs without permission. The labels are seeking up to $150,000 per infringed song, which could add up to billions in damages.
Both Suno and Udio claim that the use of copyrighted material falls under the doctrine of fair use. Universal Music and Warner Music have been negotiating AI licensing agreements with these companies and Google.
OpenAI takes some precautions with copyrighted material. ChatGPT does not share the full lyrics of certain songs, but instead offers to summarize them. After casting Sora, altman said the company would give copyright holders “more granular control” over character generation and would share revenue from video generation with rights holders.
#OpenAI #building #music #generator #Juilliard #students
