The three-part podcast Fascination touched on AI (artificial intelligence) going musical. Allison Parshall, a science journalist, multimedia editor, and podcast host, spoke to musicians about their thoughts on AI possibly taking over their industry.
Consequences of AI-Generated Music
Parshall noted that the new world of artificial intelligence had reached the music industry. We have been serenaded by AI-generated music from Google's latest model, music MusicLM.
Sofía Oriana Infante, one of the members of PAMP! and second-place entry to the 2022 AI Song Contest is a composer. She got interested in AI while studying music composition for film.
Looking back, she remembered her classmates using software to compose music. She was amazed because it was like pressing some keys on the keyboard, which worked automatically, Scientific American reported.
It prompted several questions in her head, and she pursued Ph.D. in musicology to study AI as a composition tool.
She admitted that she was furious with AI and those behind it when she reflected on what they were doing with music. However, she also realized there are two sides to it - one can use it to replace composers to earn more money, or she can use it for her craft. She decided to do something in symbiosis.
Oriana Infante and her team used AI tools to transform vocals and sound like a hurdy-gurdy or a traditional cranked string instrument. She realized that with AI, she could accomplish things humans can't easily do to improve her compositions.
Oriana Infante doesn't think AI will leave musicians jobless. Instead, she believed that it could help them improve what they do.
How Can AI Help Musicians?
For Oriana Infante, the most important thing about using artificial intelligence in compositions is doing things humans cannot do. According to her, one can ask AI to create a song while watching a sunset with your significant other. AI contributes by providing them with fresh ideas they otherwise would not have.
Christine McLeavey, the pianist and AI engineer from OpenAI, echoed the same sentiment. According to her, using AI for music opened her mind to many more creative possibilities. She uses MuseNet, a deep neural network that can generate 4-minute musical compositions with ten different instruments and combine various styles, including Mozart and the Beatles. She would feed into it the beginning of a piano piece she has played a million times. In her mind, there was only one way to play the piano piece, but MuseNet would generate tons of samples; some of them are great, and some are terrible.
She added that if she writes a piece of music and finds it's terrible, instead of thinking she's not good at it, she could console herself by writing more until she finds one that will satisfy her. She finds that mindset exciting to the human creative process.
Shelly Palmer, who spent much of his career composing commercial music for brands such as Meow Mix and Bojangles Fried Chicken, admitted that with AI, commercial music buyers would be paying less, and commercial music sellers would be earning less. However, suppose the composers and producers will use AI to make them ten times more productive and their craft ten times more powerful. In that case, he considers it a "productivity enhancers."
Check out more news and information on Technology in Science Times.
