Uploaded June 2026 | Updated September 2026, 16 minutes ago
This video features renowned keyboardist Jordan Rudess demonstrating a live musical performance with the MIT Jam Bot at the 2026 Berklee AI Music Summit. The Jam Bot, a winner of the 2025 MIDI Innovation Award, is an AI-driven system designed for real-time musical improvisation.
Key Highlights of the Performance & Collaboration:
Human-AI Interaction (0:00 - 1:11): Rudess explains how he trained the model using his own MIDI performance data, playing "freely but methodically" to ensure the AI's responses would remain musical. He describes the process as a "call and response" system where he prompts the AI and then lifts his hands to let it take the lead.
Technical & Visual Integration (0:37 - 0:52): The system, programmed by Perry Naseck, incorporates interactive elements to signal when the AI is active. While previous performances at MIT used kinetic sculptures, this version features reactive lighting to visually distinguish between Rudess's playing and the AI's contributions.
Musical Improvisation (1:12 - 5:09): The remainder of the video showcases the live demonstration of the Jam Bot interacting with Rudess in a dynamic, improvised setting.
About the MIT Jam Bot:
Developed at the MIT Media Lab (Responsive Environments group) by a team including Lancelot Blanchard, Professor Joseph Paradiso, and Perry Naseck, the Jam Bot utilizes transformer-based music language models. It is designed to act as a conversational partner rather than a static accompaniment tool, focusing on low-latency, responsive AI that blends seamlessly with live human performance.
This video features renowned keyboardist Jordan Rudess demonstrating a live musical performance with the MIT Jam Bot at the 2026 Berklee AI Music Summit. The Jam Bot, a winner of the 2025 MIDI Innovation Award, is an AI-driven system designed for real-time musical improvisation.
Key Highlights of the Performance & Collaboration:
Human-AI Interaction (0:00 - 1:11): Rudess explains how he trained the model using his own MIDI performance data, playing "freely but methodically" to ensure the AI's responses would remain musical. He describes the process as a "call and response" system where he prompts the AI and then lifts his hands to let it take the lead.
Technical & Visual Integration (0:37 - 0:52): The system, programmed by Perry Naseck, incorporates interactive elements to signal when the AI is active. While previous performances at MIT used kinetic sculptures, this version features reactive lighting to visually distinguish between Rudess's playing and the AI's contributions.
Musical Improvisation (1:12 - 5:09): The remainder of the video showcases the live demonstration of the Jam Bot interacting with Rudess in a dynamic, improvised setting.
About the MIT Jam Bot:
Developed at the MIT Media Lab (Responsive Environments group) by a team including Lancelot Blanchard, Professor Joseph Paradiso, and Perry Naseck, the Jam Bot utilizes transformer-based music language models. It is designed to act as a conversational partner rather than a static accompaniment tool, focusing on low-latency, responsive AI that blends seamlessly with live human performance.
