A University of Chicago study on mass-produced AI music raises a much bigger question for the music industry: in an age of infinite content, could artist development and human curation become more valuable than ever?
For years, parts of the music industry have gradually been replacing artist development with data, algorithms, formulas and, now, the possibility of producing thousands of songs through Artificial Intelligence. The implicit promise seems to be: if we produce enough music, statistically, some of it will eventually become a hit.
This study conducted by researchers at the University of Chicago reveals the limitations of that logic. The overwhelming majority of the AI-generated music analyzed shows extremely low levels of engagement. Even more interestingly, when one of these tracks does gain some traction, it demonstrates a significantly lower ability to carry that success over to subsequent releases. This reminds us of something that has always been at the heart of A&R: audiences do not connect with songs alone. They connect with artists, stories, identities, emotions and cultural worlds.
True A&R has never simply been about finding a song that “sounds like a hit.” It is about identifying artistic personality, repertoire, interpretation, originality and, above all, something that cannot easily be replicated.
Perhaps we are reaching a point where the infinite abundance of music makes human curation even more important. When anyone — or any machine — can produce thousands of tracks, producing more is no longer a competitive advantage. The difference once again lies in having something to say.
This does not mean rejecting Artificial Intelligence. AI can be an extraordinary creative tool in the hands of musicians, songwriters and producers. The study itself makes an important distinction between the legitimate use of AI as a creative tool and the mass production of AI-generated music.
For me, the message is different: perhaps it is time to recover something the industry has gradually begun to lose — developing real artists again. Finding people with personality, talent, repertoire and artistic potential; encouraging them to interact and collaborate with great songwriters, musicians and producers; allowing them to experiment, make mistakes, discover their own sound and build an identity.
Fewer songs designed to feed algorithms. More songs capable of moving people.
Perhaps the future of music, paradoxically, lies in recovering what has always made music meaningful: the artisanal, spontaneous, imperfect and profoundly human process of creating something capable of connecting with another human being.
The Music2Deal Blog 