AI and the Music Industry: Drake, Grimes and Google's Grand Plan

It is not an easy feeling, as a music student, knowing that our jobs could be automated. Many of us chose this profession thinking the arts were sacred territory, the last industry to become the work of machines. But as we have seen, recent developments could prove otherwise, and a deal between Google and Universal Music Group for generating music will have major implications. Is this the final death of authenticity in music, or the beginning of an entirely new creative industry? The benefits of the deal are yet to be determined, but to understand we need some context…

AI has been lurking in the background of the music industry for a few years now. Mostly in the form of niche audio production tools for mastering or separating stems. Generative AI wasn’t convincing enough until the anonymous producer Ghostwriter977 took the stage. In April they released an AI assisted track ‘Heart on my Sleeve’ that was presented as a collaboration between Drake and The Weeknd. Not only was the song convincing, but many considered it enjoyable, heralding a new age of music technology. The success of the tune provoked a new type of copyright dispute, mostly in the form of asking platforms to restrict people from being able to train models. International copyright law regarding AI is still in its infancy.

Not long after, Grimes presented a new take on the issue. In contrast to Drake’s obvious disdain to the method, she offered a 50/50 royalty split to anybody who achieved success with a song that uses an AI generated version of her voice. She stated her lack of record label or other legal bindings, essentially opening up the floodgates for AI music with her voice as the figurehead. Within a few days she retracted the openness of her offer, stating that it wasn’t permission to make terrible music with her voice.

Imitation is not yet regulated enough to provide reasonable defense for artists. Recently, the popularity of a Johnny Cash version of ‘Barbie Girl’ has opened the door for resurrecting personalities that were thought to be long gone. Reviving dead artists raises questions about authenticity. Is it worth reviving the legends considering the risk of losing what made each of them unique?

There are varying perspectives on the ethics of generative AI in the arts. There is a similar discussion happening in the world of visual art, but many people see it as the future. A recent court ruling claimed that AI generated content images cannot be copyrighted due to the absence of “a guided human hand”. Meaning the artist is lacking “ultimate creative control”, though the degree of human intervention required is still subject to debate. Some lawyers believe that existing copyright laws can hold up under the AI revolution, but the definition of human input is becoming increasingly blurred. Current laws favor traditional art and lack support for newer forms.

In a controversial move, Google has made a lunge for the market. They have signed a deal with Universal Music Group, one of the ‘Big Three’ record labels, to facilitate AI generated content with licenced likenesses. They have started an incubator with a number of people such as Ryan Tedder, Anitta and the Estate of Frank Sinatra. Regardless of your stance on this, we will soon see a wave of this type of content.

This is uncharted territory, making predictions difficult. It is comparable to when streaming services became widely used in an effort to make a return in the age of digital pirating. People will be using the technology, prohibition will only drive it underground. This is an effort to get ahead and be the first to make a profit. Better the devil you know. Though it is important that artists have control of their likeness, while also benefiting from the profits.

What does this mean for students in the industry? The bar will be raised for what is considered a “skilled” position as aspects of music production are automated. Though this also allows more people to be able to use the technology. Some current examples of this happening are AI mastering tools and Digital Production software. The technology is more affordable and accessible than before, these days anybody with a laptop has the capacity for high level sound design. This will only get better, but then what skills are still important? As always, the ability to adapt to the technology and develop a global understanding of music. Without the weight of time consuming tasks, artists can expand into other forms of media, raising the standard of music releases. Or we can just focus on the quality of the music, and make more of it.

In a world saturated with generated music we are likely to see new value attached to live music. There will always be those that ask not just ‘what?’ but ‘who?’ and ‘how?’. Performance skills will still be valuable, though the size of the potential audience is yet to be determined. There is a certain human component of music that AI is unlikely to be able to replicate. Even if it were, people may not accept the answers to the ‘who?’ and ‘how?’. People resonate with artists because of similarities in their character and experiences, and this won’t be as effective for an artist that doesn’t share a likeness with the audience.

The thought of infinite content is daunting, though the opportunities for innovation are expansive. Maybe albums in the future will be interactive, letting us change the style of a song or bring in other artists. We could all have our own personal soundtracks that are reflective of our mood, environment or preferences, constantly changing as we go through life. Artists could collaborate with fans on a global scale, with whole crowds making music together under their oversight. The possibilities are endless, we can only hope it is not overridden by profit and greed. At the end of the day, it is just another tool, only the wielder is to blame.