At The Volatile Crossing Of Formal Music And AI, Machover Vision…

Tod Machover unveiled a new concerto on October 24 at Boston’s Symphony Hall, where the Boston Symphony Orchestra performed a work that integrates artificial intelligence with live orchestral music. This piece, titled already and not yet, challenges traditional boundaries by positioning AI as an active participant alongside the orchestra rather than a mere effect or background element.

The concerto’s innovative structure involves three continuous movements where the orchestra performs acoustic instruments that trigger AI responses in real time. The AI, developed and managed by MIT’s Manaswi Mishra, listens to the musicians’ improvisations and returns its own harmonies, counterpoint, and sound enhancements, creating a unique performance during each rendition.

Machover’s ongoing experimentation with blending electronic and acoustic sounds stems from his long career at MIT’s Media Lab and his early experiences with Pierre Boulez’s IRCAM electronic ensemble. His history of integrating taped sounds, robots, and invented instruments into operatic and orchestral works informs this latest exploration, which tests the dynamic interaction between human musicians and AI.

During the premiere, the first movement emphasized the acoustic ensemble with limited AI involvement, spotlighting contrasting brass and string textures reminiscent of Bartók’s Concerto for Orchestra, premiered in the same hall in 1944. The AI contributions initially enhanced certain brass passages but remained subtle, serving more as a sonic augmentation than a dominating presence.

The second movement increased AI activity as the software responded to improvisations performed by the orchestra’s principal players. The AI not only mimicked tonal colors but also generated original harmonies and a moving chorale effect, occasionally triggering sampled vocal recordings controlled by collaborator Ana Schon. This interplay between acoustic improvisation and AI response suggested a call-and-response format where the source of sounds grew indistinct over time.

By the third movement, AI assumed a more prominent role amid a simplified acoustic texture. A highlighted AI cadenza compiled elements recalled from earlier in the concerto, interrupting a deliberately chaotic musical tableau where multiple musicians simultaneously played differing improvisations. Despite the complexity, the orchestra maintained disciplined professionalism throughout the challenging textures.

While this concerto represents a significant step in live AI integration, it stops short of fully engaging the improvisational potential of the orchestra, as the musicians worked within limited pitch material and functioned partly as triggers for the AI. The resulting texture sometimes obscured the individual contributions of human players, reducing them to components in a broader sonic experiment.

This project continues an experimental lineage that connects early sound augmentation efforts to current explorations of machine intelligence in music. Machover and his team at MIT act as pioneers probing new artistic territories rather than definitively establishing AI as an autonomous creative agent in classical music.

The performance was part of a program conducted by Kent Nagano, who included Mendelssohn’s Hebrides Overture and Reformation Symphony alongside a new orchestral arrangement by Argentine composer Alex Nante. Nante’s work reimagined Bach’s chorale harmonization in a way that honored the original but also intensified its expressive depth, linking Bach’s legacy with Mendelssohn’s interpretation.

Nagano’s programming choice connected historical and contemporary threads, providing contrast to Machover’s forward-looking concerto. The seamless transition from Nante’s one-movement piece to Mendelssohn’s symphony underscored the continuum of musical tradition and innovation within the concert.

The inclusion of AI in a major orchestral work underlines ongoing shifts in classical music performance and creation. Machover’s approach, combining live electronics and AI with traditional instrumentation, might foreshadow future collaborations where technology supplements rather than replaces human artistry.

Despite the concerto’s ambition, the balance between AI and musicianship remains a subject for further development. The work’s improvisational framework, while offering infinite variation, also highlights challenges in integrating machine learning meaningfully with nuanced human expression during live performance.

Machover’s catalogue, enriched by this concerto, illustrates an evolving relationship between technology and classical form. As orchestras and composers increasingly explore AI’s potential, such works could influence programming choices and audience expectations across the music industry.

The Boston debut of already and not yet demonstrates both the possibilities and current limitations of AI in concert music. Its experimental nature invites ongoing dialogue about the role of technology in expanding the boundaries of composition and performance.