Navigating today’s technological landscape, it’s evident that keeping pace with the rapid metamorphosis of industries under the umbrella of Artificial Intelligence is, to say the least, demanding. While most conversations center around AI’s influence on visual arts, it’s hard to miss the crescendo building in the music industry.
Now, have you heard of Noonoouri? I didn’t. At least until a few days ago when the following Forbes’ article sneaked into my news feed.
Let’s first talk a bit about Noonoouri. She is a cartoony, 19-year-old fashionista hailing from Munich, Germany. She has worked with most of the top brands in the fashion industry and continues to wow consumers with her unique look. Noonoouri balances her platform between social good and promotion. She’s vegan, advocates for sustainable fashion, and refuses to wear furs while making countless cameos with fashion brands all around the world. At least that’s what www.virtualhumans.com says. Because Noonouri’s existence is confined to the digital realm. Noonoouri is the creation of Joerg Zuber, founder of creative agency Opium Effect.
And here is where the plot thickens. This last week, she’s also been landing a recording deal with the Warner Music Group and releasing her first single, Dominoes, which you can listen to here. The voice behind it is naturally, an AI modified rendition of a real human voice, that grants Noonoouri a unique voice. The technology that was used is generative AI – think Chat GPT but for crafting voices. How will the deal work, you might wonder? Well… Human touch remains central, with individuals crafting tunes and reaping royalties in the traditional songwriter fashion. But here’s the twist: AI is now adept at producing entire songs.
The tools are there already. OpenAI has MuseNet, which uses something similar to its Chat GPT to `generate 4-minute musical compositions with ten different instruments`. On the other hand, Google’s Magenta Studio is built on neural networks and ingests vast arrays of music, detecting recurrent patterns and producing tracks reminiscent of… human compositions? Or not really. To make an analogy with cooking, the AI is only following a recipe now, rather than inventing one. It follows a script that sometimes delivers.
Sir Paul McCartney also used `stem-separation` tools that split original digital audio files in components such as vocals and instruments. This helped him extract John Lenon’s vocals and piano from an old tape, culminating in the last Beatles offering.
The effects of this will be huge to say the least. Producers will probably remonetize dead artists, and anyone will be able to produce musical compositions (originality up for debate).
And as with any disruptive technology, there are cries for oversight and the looming shadow of job redundancies. But for me, this is not very different from what any other major scientific discovery has done to a lot of industries throughout history. Artists still have live music. And why do people like live music? Because it’s risky. Because you don’t know what you’re gonna get. Because it’s unique in terms of sound and atmosphere. And this is not replaceable. So maybe instead of fighting something that is already here and generating effects, maybe our focus should be on how to use this to make better music and better art. Remember, AI, potent as it might be, is but a tool in our vast musical ensemble.
