Jaden Ezzo

With the rise of AI generated music, consumers, educators and artists debate about the ethicality behind using it to make music.

“When AI looks at music, it tells you a little bit about you [and] how humans think about music,” Jeremy Chesman, music professor at Missouri State University, said. “Sometimes they’re using AI to kind of understand more about how people think about music, and its structure. In some ways, that is interesting, but then in other ways the arts are supposed to be a present representation of what human beings are, that’s why they’re called the humanities. If you take humans out of it, it doesn’t really have the same soul.”

Music consumers, like sophomore Thomas Jordan, said that using AI strips back the human nature of music.

“It takes the focus away from human artists who put their time, and effort and money towards having their music, [and] having their sound heard,” Jordan said. “Replacing it with something made by a machine, it’s both disingenuous to people who create music and to replace it with something that’s been made like that is kind of immoral.”

Educators, like assistant professor at the University of Michigan Julie Zhu said AI can replicate emotional depth.

“AI is pretty good at generating music that already exists, for example [Johann Sebastian] Bach or Taylor Swift,” Zhu said. “If you believe those artists have emotional depth and nuance, then yes, AI can replicate the ’emotional depth and nuance of human-created music.’ But our tastes are constantly shifting, and novelty still counts for something.”

There are several sites that aid in the creation of AI music. One of these sites is Suno AI, where an individual can type in a prompt and create an entire song with vocals and instrumentation.
Sites like this have been under legal scrutiny.

“It has been proven in the court of law that Suno AI and other music generating programs illegally used 100s of terabytes worth of music from artists that never gave their permission to have their songs used to train their programs,” Edward Stengel, founder of the band Pawns or Kings said. “As a result, these programs copy and recreate song data from these musicians without any money going to the original creators.”

Music instructor at Missouri State University Kyle Aho said the human element is the critical ingredient to creating art, whether visual, musical or written.

“It’s hard to quantify and put some sort of numerical figure on how important emotional depth is to great music making,” Aho said.

Zhu runs Deep Drawing, a lab that combines music and machine learning to visualize an image.

“We use machine learning to synthesize sounds from the drawing board and to perform surface sound source localization, which is, to understand from just the audio, what is being drawn, coordinate-wise,” Zhu said.

Deep Drawing uses humans and machines, looking how they interact for purposes of art and coding.

“Deep Drawing is, at its core, an artistic endeavor, and machine learning here is used as an investigatory tool,” Zhu said. “We aren’t solving any practical problems for society … [but] we believe that [Deep Drawing] is worthwhile.”

From weddings to funerals, Chesman said that the music that impacts individuals the most has personal meaning.

“It’s not really expressing anything [AI generated music],” Chesman said. “I think the songs that move us the most, that we remember, has meaning to us, the times that are important for life.”

With AI being used in many industries, the question of how people are going to respond to AI music comes into frame.

“It’s hard to say where people’s morality on it is going to be in the future”, Jordan said. “In the best case, the artists who use AI wouldn’t be accepted in the first place. It wouldn’t be praised at all … but it’s hard to tell where it’ll go in the future.”

A survey from the music streaming service Deezer said 40 percent of people would skip a song if they were shown it was made by AI and 45 percent of people would filter out AI music from their playlists.

AI generated music can be compared to social media. Specifically, AI generated music can create more isolation when music is supposed to foster community, similar to social media.

“We’ve seen the same [isolation] with what has happened with social media, that perhaps we did not consider approaching that with more caution at the beginning,” Chesman said. “I think the mental health of the larger society has suffered because of that. I don’t really see that AI is going to help people suffer less and be more connected to their neighbors. … You have to build community around something that really unites people.”

Aho said human made music creates a sense of imperfect perfection, but AI threatens this by invoking sounds that are impossible for humans to replicate.

“We expect singers to sing in a way that is not real, we expect drummers to play in a way that isn’t real and pianists and so on,” Aho said. “This [AI] just takes things to a completely other level. The fear is that consumers and young people would almost begin to prefer AI [generated] music over human music.”

With the usage of apps like Garage Band and Band Labs, Jordan said people who want to create music can use these apps instead of AI.

“People who make it [AI generated music] often say that either they’re not born with enough talent, which I disagree with, or they do it because they feel like they can’t,” Jordan said. “I don’t think that’s true at all, anybody can make music.”