A bunch of 200 musicians signed an open letter urging tech firms and developers to not undermine human creativity with AI music generation tools.
The list of signing artists is so extensive and diverse that it could make for a terrific Coachella lineup – including Billie Eilish, the Bob Marley successor, Chappell Roan, Elvis Costello, Greta Van Fleet, Imagine Dragons and Jon Bon Jovi, the Jonas Brothers, Kacey Musgraves, Katy Perry, Mac DeMarco, Miranda Lambert, Mumford & Sons, Nicki Minaj, Noah Kahan, Pearl Jam, Sheryl Crow and Zayn Malik, amongst others.
“If used irresponsibly, AI poses an amazing threat to our ability to guard our privacy, our identity, our music and our livelihoods,” the letter said. “Some of the biggest and strongest firms are using our work without permission to coach AI models. … This could be a disaster for a lot of working musicians, artists and songwriters just attempting to make ends meet.”
These artists are right. The AI models that generate recent music, artwork and lyrics work by training on massive datasets of existing work, and usually asking you to remove your work from these models is an exercise in futility. It could be like considered one of these artists trying to forestall anyone from pirating their music – that's just not realistic. It's already possible to create convincing deepfakes of popular artists, and the technology is just recovering.
Some firms like Adobe and Stability AI are working on AI music generators that use licensed or royalty-free music. But even these tools could negatively impact artists creating scores for television commercials or other beats that an artist might license for his or her work.
Historically, musicians have all the time lost out as technology has develop into more sophisticated. First, it was file sharing that made it easy to get free music; Streaming proved to be an answer to this problem, however it shouldn’t be an answer that satisfies artists. The Union of Musicians and Allied Workers (UMAW) has been working for years to secure higher streaming payouts for artists – the guild's artists estimate that Spotify's average streaming royalty rate is around 50% $0.0038, or a few quarter of a cent. So it is sensible that musicians remain skeptical of this recent technology.
Authors have also spoken out against the rise of generative AI. In July, over 15,000 authors – including James Patterson, Michael Chabon, Suzanne Collins, Roxane Gay and others – signed an identical open letter addressed to the CEOs of OpenAI, Alphabet, Meta, Stability AI, IBM and Microsoft.
“These technologies mimic our language, stories, style and concepts and produce them back to life. “Millions of copyrighted books, articles, essays and poems provide the 'food' for AI systems, countless meals for which there isn’t a bill,” say the authors. letter reads.
But these tech firms aren't listening. You can still go to ChatGPT and ask it to supply a Margaret Atwood-style passage – that's not necessarily good, however it does indicate that the nice language model has absorbed “The Handmaid's Tale” and may spit out a degraded version of it It. And since copyright law isn't necessarily sophisticated enough to cope with generative AI, legal recourse is pretty useless at this point.
“This attack on human creativity should be stopped,” the musicians said of their letter. “We must protect ourselves from the predatory use of AI to steal the voices and likenesses of skilled artists, violate creators’ rights, and destroy the music ecosystem.”
