AI and the Composers of Classical Music

Written by GPT-4 using the ChatGPT GUI and edited by Dr. Ifan Payne. For information about AI and the opera, click here to read Bernadette Snider’s blog article, “A New Odyssey.”

This is the first of two articles on Artificial Intelligence in music which were written by Artificial Intelligence software GPT-4 , a Large Learning Model (LLM) using the ChatGPT bot and edited by me. Can you tell what was written by the AI bot and what was written by me?

The accompanying illustrations were selected from images that I generated from the AI programs, DreamStudio, Midjourney and Dall-E.

Artificial Intelligence (AI) is rapidly becoming all pervasive in the creative arts, not least in music. Composers of classical music can use AI in various ways to aid their creative process.

1. Composition assistance: Composers can employ AI algorithms to generate musical ideas, harmonies, melodies, or even entire compositions. By training AI models on a vast dataset of classical music, composers can leverage these models to generate novel musical material that adheres to specific stylistic rules or follows certain compositional patterns.

2. Harmonic and melodic analysis: AI can be used to analyze and extract patterns from existing classical music compositions. Composers can feed large amounts of music into AI algorithms, which can then identify common harmonic progressions, melodic motifs, or structural elements. This analysis can provide insights into compositional techniques used by classical composers, helping contemporary composers gain inspiration or develop new ideas.

3. Orchestration guidance: Composers can use AI tools to assist them in orchestration, helping them determine the most appropriate instrumentation and voicing for their compositions. AI algorithms can simulate different instrumental combinations and suggest orchestration choices based on the desired musical effect or the composer's specific requirements.

4. Generative models: AI-powered generative models, such as Variational Autoencoders (VAEs) or Generative Adversarial Networks (GANs), can create new musical compositions based on learned patterns from existing classical music. Composers can guide the generation process by providing input or constraints, and the AI model can produce unique compositions that align with the style or characteristics of classical music.

5. Music recommendation: AI can assist composers by recommending existing classical compositions that align with their preferences or specific requirements. By analyzing a composer's musical preferences, AI algorithms can suggest compositions from a vast repertoire that composers may not have been previously exposed to, helping them discover new ideas or drawing inspiration from diverse sources.

6. Real-time feedback and improvisation: AI systems can provide real-time feedback to composers during the compositional process. Composers can interact with AI models that analyze their musical input and provide suggestions for development, improvement, or alternative directions. This interactive feedback loop can facilitate experimentation and exploration of different compositional possibilities.

Here are a few examples of classical musical compositions that have utilized AI:

1. "AIVA" by Pierre Barreau: AIVA (Artificial Intelligence Virtual Artist) is an AI composer created by Pierre Barreau. AIVA has composed original classical music pieces that have been performed by orchestras around the world. The compositions created by AIVA demonstrate the AI's ability to generate complex and emotive musical compositions in the style of classical music.

2. "Hello World!" by Kostas Vomvolos: This composition was created using the "DeepBach" AI system developed by researchers at Sony CSL. DeepBach is an AI model trained on the works of Johann Sebastian Bach. "Hello World!" is a choral piece that blends Bach's style with contemporary elements, resulting in a unique fusion of classical and modern aesthetics.

3. "Iamus" by Francisco Vico, Gustavo Diaz-Jerez, and their team: Iamus is an AI system developed by researchers from the University of Malaga. It has composed full orchestral works that have been performed by professional ensembles. These compositions showcase the AI's ability to create intricate and expressive music, combining traditional and innovative approaches.

4. "Uncanny Valley" by Anna Meredith: Anna Meredith, a British composer, collaborated with the AI technology company Jukedeck to create "Uncanny Valley." This composition was part of a project exploring the intersection of AI and music. Meredith used AI-generated material as a starting point and further developed it into a complete orchestral piece.

5. "The Next Rembrandt" by Jeroen van der Meij and his team: Although not a musical composition, this project utilized AI to create a painting in the style of Rembrandt. The team used machine learning algorithms to analyze Rembrandt's style, techniques, and subject matter, resulting in a unique artwork that mimicked the master painter's aesthetic.

This OperaBlog was contributed by Ifan Payne, Member At-Large, Santa Fe Opera Guild Board of Directors, June 1, 2023.