Why AI Music Won’t Make You a Musician
AI music models are tools like any other. There are no shortcuts nor limitations to creativity.
While Eurovision has not always been noted for its musical quality, the competition has always been a great distiller of musical trends and a good illustration of the general musicpolitik across the continent (and a handful of places from elsewhere). This year’s contest is no different.
Already, it breaks precedent by taking place not within the borders of last year’s winner but instead within the realms of the runner up here in the UK — who had its speedy reaction to the war in Ukraine rewarded with second place in last year’s competition (although the track was a bit of a belter too).
But it isn’t only geographically that Eurovision 2023 breaks precedence. This year is the first time in its history where not a single Eurovision entry will include a key change. Compare this to ten years ago when a third of participants entered into the competition with a key change in their last chorus, or consider that it has been 16 years since we last saw a winner with a key change — and it is clear that something has changed (if not the keys).
For those who need a recap: a key change is exactly what it sounds like — a shift in a song’s tonality, often (but not always) towards the end of the song (think the end of Living on a Prayer by Bon Jovi or in the chorus of About a Girl by Nirvana).
This development (or deterioration depending on your views — but more on that later) is not just a European one. Chris Dalla Riva, a musician and data analyst at Audiomack, found that about a quarter of number-one hits from the 1960s to the 1990s included a key change. But from 2010 to 2020, there was just one top song: Travis Scott's 2018 track, Sicko Mode.
Now, this may seem like a sad development of sorts. But now ask yourself, honestly, had you noticed? Do you expect your Eurovision experience to be diminished this year because of it? Or do you, like most, believe that Eurovision has been on an upward trajectory over the past years, both in song quality and in general enjoyment?
The answer is probably the latter. And I write about the loss of the key change not because I do not lament it but to show that ultimately the key change is not a good in and of itself. Key changes can elevate a song to legendary status or dishonestly prop up a song to irritatingly cheesy new heights. In short, key changes are not symbiotic with good music, and they can be used both for good and bad just like any other musical technique or tool.
At the same time, its loss is not without replacement either. The demise of the key change coincided with the wider adoption of computers in music and the rise of rap & hip-hop (where a key change did not matter as the lyricist would not follow it). Indeed, the evolution has continued and, nowadays, songs incorporate entire genre changes in their music — wherein the song shifts into an entirely different type of track midway through.
The same will be the case with AI music, which this week hit the spotlight when a TikTok user reportedly used AI to make a bootleg track by the uber-famous Drake. The song — which was likely used with a voice cloning tool such as Elevenlabs or Murf — has since been taken down, but not before it was listened to some 20 million times. Say what you want about the track, but I know I for one could not have made it.
This gets to the crux of my point: that just because such tools exist — and many more on the horizon such as MusicLM by Google in June — doesn’t make one an expert at using them. Whilst AI will lower the entry bar for many skills, it will not take away from those who are dedicated to their craft.
That is because, contrary to popular opinion, it is not computers and rap that made modern music sound the same. More likely, it is because most of all pop music from the past 20 years or so has been written by the same guy: Max Martin. To date, Martin has written or co-written 25 Billboard Hot 100 number-one singles, with him also producing, or at least co-producing the majority. With more people making more music, it is much more likely that we will see more divergent art forms emerge.
In fact, analysis from JP Morgan shows that 70 per cent of listening on streaming platforms is catalogue (AKA old music). The industry is beyond ripe for disruption, and even if we did not know it we are all voting with our ears and yearning for something different. With AI music, this art form will be further democratised and, just like in the early days of social media, we will see a boom in creativity and accessibility as a result.
Here at ACQUAINTED, we take this view with all AI tools. Stagnated creative monopolies are going to be broken as more people seize the means of creative production and once again create culture from the ground up. Regardless of your wider views on AI this can only be a good thing in terms of creative output.
In this sense, just as the loss of the key change did not lead to the end of Eurovision, AI music will not be the end of music, nor will it be the end of the musician. Instead, it will create more music — both good and bad — and make more musicians — both good and bad. These are tools that will need practise in order to be used effectively, require an intimate understanding of both their strengths and weaknesses in order to harness properly, and — as with anything else that requires skill — require plenty of time to hone.
That is why the existence of AI music won’t make you a musician — and is why if your billion-dollar empire monotone voice can be replicated by a random TikTok user then maybe it’s time for a genre change.
If you would like to learn more about how music is changing and all about the Generative AI tools that are available today then get in touch at: elliot@acquainted.studio.
