New York, NY (Top40 Charts) It's hard to look at any musical, visual or graphical work lately without having a small nagging doubt at the back of your mind: "Was that made by a human or a machine?". For the moment at least, it appears that AI is here to stay and is altering the creative world in unprecedented ways.
Many see it as the next logical evolution in the creative arts while just as many decry the potential for soulless, mass-produced content that could potentially remove humans entirely from the process. In terms of music, the technology is raising all manner of questions legally and ethically and recent cases are showing that the issue is far more complex than many ever suspected.
The Current State of AI Music
The last year or two has seen a surge of AI-generated music, whether that means entirely new compositions or those that take more than a little from the current stars of the world. One of the biggest trends in AI music is taking other songs, unrelated to an artist, and then feeding it into an AI program using that artist's voice as a sample. What comes out is a cover song that the famous artist never actually did but is more than capable of fooling a casual listener.
Major names like Drake have already started moving into legal battles to try and stop the practice, but as with any new technology, the law is taking a painfully long time to catch up to the reality of the situation. As it stands, even copyright doesn't have much chance of stemming the flow of AI covers with attempts by labels and major companies to have songs taken down being hit or miss at best.
What has surprised many is how effective the process is. These AI covers currently take minimal work to produce quality that sounds almost spot-on to the original artist's voice with almost no editing required. More impressively, it isn't just those artists with more 'standard' voices that are at risk.
We recently dropped the news that the unmistakable Sean Paul has a new song out with Reggae artist Beres Hammond, and it's hard to imagine any AI system managing to imitate either one successfully. However, thanks to the unique way that the software learns, there are already projects out there coming close with only a few months of effort.
The Impact Going Forward
There are a growing number of voices out there claiming that this is just the beginning of an AI takeover of music, and soon the whole industry will just be machine-based. The reality is, as always, a lot more complex.
We can go back to Drake for an example. Yes, AI has proven capable of copying the man's voice, but that's all it can do: AI cannot become the person. A top-tier artist is far more than just their music, being the sum of their abilities, their personalities, their charisma and a thousand other factors.
Drake, for one, is famous for traits beyond just his music. The Canadian artist has been known to drop bets in casinos, a unique quality that adds something relatable to his public image. Canada has some of the best online casinos that's popular and easily accessible to a range of people who can log on remotely. AI might be able to emulate a style or pick up on a catchy chord progression, but people are drawn to artists that share their hobbies and the added connection that brings; that's something AI does not have a handle on.
There is also the issue of having a real person on hand to engage with fans. Some point to acts like Gorillaz or the recent 'vocaloid' trend as examples of artists that don't really exist, but in each case there is still a real human behind the music. While AI chat software is currently able to emulate a standard human conversation, it is still a lot way off from emulating a full personality.
So Is AI the End to Human Music?
It is important to remember that AI has one critical flaw that humans simply don't have. By nature, AI can only work based on material that already exists, and while it can generate 'new' material, that material will always be derivative of something else. Humans retain the distinction of being creative and able to create entirely new concepts.
Are we going to be seeing more AI music? Almost certainly, given the low cost and decent quality possible. Will it be a complete replacement of our top artists? Almost impossible.
