New AI agents MACAT and MACataRT redefine music co-creation
Two AI systems MACAT and MACataRT turn musicians into real-time collaborators with AI
Deep Dive
Keon Ju M. Lee and Philippe Pasquier introduce MACAT and MACataRT, two musical agent systems designed to support music performance and improvisation in co-creative spaces. MACAT is optimized for agent-led performance using real-time synthesis and self-listening, while MACataRT offers a flexible environment for collaborative improvisation through audio mosaicing and sequence-based learning. Both systems emphasize training on personalized small datasets, fostering ethical and transparent AI engagement that respects artistic integrity.
Key Points
- MACAT and MACataRT are AI agent systems from Simon Fraser University designed for real-time music collaboration
- MACAT autonomously leads performances using self-listening synthesis, while MACataRT enables improvisation via audio mosaicing
- Both systems train on small personalized datasets to prioritize artistic integrity and ethical AI use
Why It Matters
These AI agents empower musicians to explore uncharted creative territories while keeping human expression central.