Entertainment News

AI came for music. Until now, there is a big ocean between humans and algorithms


The production is polished, the drums are on point and the vocals sound good, but a titan of the Newfoundland and Labrador music scene hears something off about “It Could Be Worse” and “Tales of The Atlantic”, two songs generated in less than a minute by a powerful algorithm.

“He’s a country singer, so it’s wrong. And the lyrics don’t really rhyme,” said Bob Hallett, a founding member of the Newfoundland folk-rock group Great Big Sea. “It just seems a little strange.”

Hallett had just finished listening to the catchy tunes, created using a generative artificial intelligence tool called Suno, using prompts capable of describing any Great Big Sea jam: Celtic, folk, live , keen.

They didn’t accomplish their goal, Hallett said. On a scale of one to ten, 10 being the band’s hit “Ordinary Day”, he gave the songs a two.

But experts say that rating could rise quickly. Technology like Suno’s is advancing rapidly and its results are only getting better, said Jimmy Lin, professor and director of the University of Waterloo’s artificial intelligence institute.

Suno is one of several companies creating generative artificial intelligence software that allows users to create original songs using text prompts. Users can create instrumental tracks or songs with lyrics, which can be generated by the program or provided by the user. But if the user comes up with copyrighted lyrics — the opening lines of Nirvana’s “Smells Like Teen Spirit,” for example — the program won’t create the song.

Suno also does not create songs intended to resemble the work of other artists, such as “Heart on My Sleeve,” the AI-generated song using unauthorized similar vocals from Drake and the Weeknd that caused controversy in the music industry last year. . When asked to do “a Great Big Sea song about cod fishing”, Suno’s resulting tunes had a pensive, Celtic feel, but they sounded nothing like an authentic piece by the band .

Google is working on similar software, called MusicFX, which can be sampled through its AI Test Kitchen site. And Adobe last month unveiled Project Music GenAI Control, which it described as an “early-stage AI generative music generation and editing tool.”

In December, Microsoft introduced a Suno-powered song generator for its Copilot chatbot, a program that uses artificial intelligence to simulate a conversation with users.

The technology behind these programs is similar to that which powers the ChatGPT chatbot.

Lin said such programs use massive data sets to “train” algorithms, or step-by-step processes, to take any starting point and predict what the next one will be. So while chatbots trained with text can predict the next word in a written response, a music generation program is trained using sound to predict the next “acoustic sequence,” he explained.

The New York Times sued Microsoft and OpenAI, owner of ChatGPT, in December for using its stories to train programs.

Lin said the companies behind AI platforms that create music could find themselves in similar trouble if they had trained their algorithms with the work of artists who had not given consent or received compensation for their music being used in this way.

“Whether this is fair use or not remains an unresolved question,” he said in an interview. “It will sort itself out. The court always does that.”

Suno’s website does not indicate what data it used to build its program, and the company did not respond to a request for comment.

Hallett said he wouldn’t be surprised if the algorithms learned from a few Newfoundland bands. He said the songs produced had some signature sonic trademarks, including tight melodies and heavy strumming on acoustic guitars, which he and his fellow producers cultivated over the years recording albums for groups like the Ennis Sisters, Shanneyganock and The Fables.

He wasn’t bothered by the music used to form these programs, however, pointing out that artists have long faced platforms like YouTube and Spotify that already significantly reduce musicians’ income.

“There’s a little sense of abandonment in all of this…it’s so hard to control the situation,” Hallett said. “Creative work is really about driving sales at your shows or finding commercial placements. Even at the highest levels, people don’t really make money selling records anymore.”

Lin said the AI-generated songs would likely be used by advertisers who need a catchy jingle for a commercial. And he thinks it could start soon given the breakneck pace at which these tools are evolving.

“We’re not talking about years or decades. We’re talking about months,” Lin said.

But Hallett said anyone looking to use music to connect with an audience would be better served by human beings.

“It’s easy to be afraid of AI,” he said. “But we’re all drawn to sincerity in music. We want to hear people who tell us a real story and convey real emotions. And the computer just can’t do that.”

This report by The Canadian Press was first published March 3, 2024.

With files from the Associated Press


Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button