Last time we explored one of the advantages of AI and brand music. Now I’d like to talk about one of its disadvantages.

Actually it’s more like a disadvantage of digital music in general.

Digital technology has opened up a whole world of musical accessibility and flexibility. For instance, we have amp modelling software. Back in the day, your guitar sounds were limited by hardware. If you didn’t have the amp, you didn’t have its sound. These days you can plug your guitar into your computer and access a seemingly endless array of amp sounds. And a lot of them are quite good.

This illustrates both digital tech’s greatest strength and its greatest weakness. Most digital technology is used to recreate things that already exist in the real world. I can get my guitar to sound like it’s coming through virtually any amp, but there’s currently no plugin to make it sound like something we’ve never heard before.

With AI, we take this paradigm a step further. When creating AI music, the AI harvests existing or inputted media, cannibalizes relevant pieces, and presents something “new”. This allows us great flexibility, but it can also put us in a vicious cycle where everything is a slightly different version of something that came before. It leaves little room for genuine innovation.

Now for a potentially unpopular opinion:

The future of music is analog.

The latter half of the twentieth century was replete with musical innovation, and it all came from the analog realm. This took us from one musical paradigm shift to another. The seventies sounded very different from the sixties, the eighties sounded very different from the seventies, and so on.

For instance, people figured out that if you drove a guitar amp a bit harder, you would get distortion. Depending on how hard you drove it, and how you played, that sound was actually pleasing and musical. This essentially gave birth to hard rock, heavy metal, punk, and countless other genres and sub-genres.

The synthesizer was first invented in the 50s, but didn’t really come into its own until the 80s. These were sounds no one had ever heard before, and it took music to places it wouldn’t have gone with traditional instruments.

But as things have become more digital, that type of innovation has stagnated. There have certainly been shifts in style, but very little in the way of innovative sound.

If you want to produce something using existing elements, current AI will do nicely. If what we want is genuine innovation, we’re going to have to think a bit more broadly.

I’m certainly not against AI. Ultimately, AI is a tool. Any tool’s validity lies in its utility. Digital tech brings us a wealth of flexibility, but I feel that genuine innovation is going to come from clever people working in the real world. If we want to put music on autopilot, AI will do fine. If we want music to be exciting and innovative, it will have to come from people.