Will Artificial Intelligence Replace Musicians?

With technology constantly evolving and the leaps that artificial intelligence has taken, it’s not uncommon for many of us, across professions, to wonder whether A.I. will ever be able to replace humans. This is a very interesting subject, because technology has facilitated the careers of musicians by leaps and bounds over the past century. Not long ago, it was difficult for musicians to record their music, and now we have talented performers recording themselves in their bedrooms. So, naturally, this makes me wonder what’s next for classical music and what role artificial intelligence will play in this industry. 

Recently, we’ve seen artificial intelligence create a new Beatles song. In fact, artificial intelligence has been taken as far as to complete unfinished compositions by great minds like Beethoven. For classical performers, what does this mean for us?

My understanding of artificial intelligence is that it takes an incredible amount of information and turns it into analytic parameters. Artificial intelligence can process a jaw-dropping amount of information really quickly. If all that analytic power is directed to the history of music, evidently, it’ll be able to recreate compositional patterns that, undoubtedly, will be pretty close to the original compositional style (for example, finishing Beethoven’s 10th Symphony).

What A.I. will never be able to recreate, however, is the spiritual and metaphysical connection that makes up a composer’s intentionality: or, in other words, the emotional part of music. For me, this is the realm of the ineffable, the world of metaphysics. It’s in this realm that we find subtlety, humanity, in the sonic space, and, while A.I. is great at processing information, it doesn’t feel or create art the way a human would. 

From the performance perspective, it’s exactly the same. Artificial intelligence may be able to replicate what a performer does, but it won’t be able to stray far from that; it won’t create anything that’s truly inventive. In a way, this is just a translation of information, in which, like a broken telephone, a part of the message will get lost: the emotionality, the humanity, the realness. Especially when it comes to a live performance, the present moment dictates everything, because it’s the moment in which the performer cultivates the emotional space they play in. I don’t believe that A.I. will ever be able to replicate that. It’s the difference between going to a nightclub or a bar and listening to the highly produced music that’s blasting through the speakers, versus taking in the magic of a live performance. It’s just not the same. We (both performers and audiences) need that emotional moment, that connection. 

Nonetheless, I think A.I. could be incredibly beneficial to the didactic process of learning to perform. At this level, A.I. could categorize the information that makes up your repertoire and your expertise, and lay out a plan for you to reach your next milestone. It could even create your repertoire, so that it’s balanced, according to your specific needs and goals. In fact, this technology is probably already being used. 

I view musical performance as a pyramid: At the base, you have technique. This is a very rational process, which requires you to truly perceive your body and understand the way your mind and your physicality interact. In other words, technique is the perfection of pure movement. In the middle of the pyramid, there’s musicality. A.I. could participate in this second step: giving you options of how to phrase a section, where to breathe, how to harmonize each element, etc. And, at the very top of the pyramid, there’s this intentionality, this ineffable quality, that I previously wrote about. At this stage, there’s almost nothing that A.I. could do, because it comes from within the human spirit. 

In the end, what separates a legendary artist from a musically proficient one is this last section of the pyramid.

Pablo Sainz Villegas