One of the first encounters I had with a music teacher was at 7 years old. Singing in a children’s choir where its director would have her own teaching methods, usually passed down to her by previous professors or university classes. It was common to see music educators adapt these methods, like Kodály (developed more than 70 years ago), and make them their own, simply because contextually and culturally, the foundations and needs of students are different.
Several years later while studying composition in college, music education was extremely structured, following a curriculum strongly based on foreign systems. So we asked ourselves, where’s our Colombian music, or how can I learn about J.S. Bach through more contemporary methods or about Nine Inch Nails’ synths and sound deconstruction, or how should I analyze a sonata if I didn’t know the historical context of its composer?
The Power of AI in Music Education
Artificial intelligence and machine learning have enabled humans to pour endless amounts of data at obtain fast results never before seen. When applied to music, many efforts for AI-composed music have been created. Analyzing thousands of melodies, harmonies, chord progressions, beats per minute and textures to create new songs or tunes.
However, AI-powered music education is at its nascent stages. Traditionally, music teachers have passed along their knowledge to their students in masterclasses. But acknowledging that music is the only intangible art, music education is a largely biased practice. Its methods and expressions have been filtered by the analysis of previous and even legendary composers, teachers and music researchers.
Beethoven once said that music will never be perfect. Why? Because there are infinite ways to interpret or perform music. There will never be an exact performance after the other. Maybe a violin will be a quarter of a tone off tune, or perhaps the dynamics have increased one decibel compared to the previous version. The possibilities are endless.
Watch the pattern: It’s all context
Machine learning allows us to analyze thousands of historic sources and cross reference its results with music, leading to the possibility of understanding music with a whole different perspective and use it as a tool for creative and unpredictable compositions.
It is no secret that music is a mirror of history. But we have depended on oral teachings and text books to understand this. Music has served as a way to express current situations from political, social and even economical circumstances. If we used AI the other way around, analyzing music and context to understand history, we would result with billions of data points to help us unravel more unknowns. From the analogies between Mozart’s ‘The Abduction from the Seraglio’ and Turkish music, where the opera didn’t only reflect Mozart’s personal moment in life, but suggests the impact of the Turks in civilization, both politically and culturally.