If you’ve been on the internet recently, you’ve probably seen more than enough bizarre Sora-generated clips clogging up your newsfeed. Now, if reports are true, even more OpenAI-generated content could soon occupy a near-unavoidable place in what you consume online.
ChatGPT-maker OpenAI is reportedly working on a generative AI music creation tool, collaborating with students from New York’s prestigious Juilliard School on the project. According to The Information, which leaked the news, it’s unknown whether the project will be a standalone tool or integrated with another OpenAI product like video generator Sora or ChatGPT.
Details were sparse about the tool’s capabilities, but The Information says it could add music to pre-existing videos, such as creating a soundtrack for a short clip or an instrumental for an existing vocal track. OpenAI isn’t a total newcomer to the music game. It actually released a now-discontinued music generation tool, Jukebox, in 2020, allowing users to generate AI music from preset genres like reggae and blues.
Many in the music industry might not be too pleased about the AI giant’s entry into the market. Since last year, there’s been a flood of AI-generated content hitting streaming services like Spotify. Some AI bands, such as The Velvet Sundown, have managed to dupe unsuspecting listeners into believing they’re real, racking up millions of views on sites like YouTube in the process. A lyrically questionable parody song created with generative AI tool Udio even reached number three on Spotify’s viral song charts earlier this year.
Some figures within the music industry have already spoken out about how AI giants are failing to compensate creators properly, with music legend Paul McCartney urging governments to legislate to ensure human musicians get their due. Meanwhile, fraudsters have used AI music generators to con platforms like Spotify out of streaming revenue.