THUNDERBOOM

MUSIC TECH

INNOVATION LAB

THUNDERBOOM

MUSIC TECH

INNOVATION LAB

THUNDERBOOM

MUSIC TECH

INNOVATION LAB

THUNDERBOOM

MUSIC TECH

INNOVATION LAB

THUNDERBOOM

MUSIC TECH

INNOVATION LAB

THUNDERBOOM

MUSIC TECH

INNOVATION LAB

THUNDERBOOM

MUSIC TECH

INNOVATION LAB

Jeroen Ermens vs AI: Signal Interference

1.1.2025

Open Culture Tech

Jeroen Ermens is a Hammond player exploring ways to innovate his antique instrument using new technology. With his project Signal Interference, he searches for methods to jam as a solo artist with his own technological sparring partner. Within the framework of Open Culture Tech, Jeroen collaborates with Bureau Moeilijke Dingen to develop a performance tool that combines real-time music, machine learning, and visuals. The result is a musical battle between Jeroen and AI.

Jeroen Ermens is a Hammond player exploring ways to innovate his antique instrument using new technology. With his project Signal Interference, he searches for methods to jam as a solo artist with his own technological sparring partner. Within the framework of Open Culture Tech, Jeroen collaborates with Bureau Moeilijke Dingen to develop a performance tool that combines real-time music, machine learning, and visuals. The result is a musical battle between Jeroen and AI.

 

The idea stems from Jeroen’s love for Hammond battles, such as those performed by Orgelvreten. The growing number of musical AI tools offers new possibilities to engage in battles with machines. However, this is often only possible using tools from large tech companies, which typically do not handle your data and privacy responsibly.

 

 

Together with Bureau Moeilijke Dingen, Jeroen has developed an application that listens to his playing, analyzes the data, and compares it to a database of tracks that he himself has recorded. The system can then generate a live response to Jeroen’s playing in his own style. The algorithm also introduces surprises, creating a sparring partner that not only matches him but can also catch him off guard.

 

The system is based on an old concept: a Markov Chain. This is a mathematical model that describes a sequence of events where the probability of each next event depends only on the current state, not on the events that came before. In other words, the system “remembers” only where it is now and uses that information to determine the next step. The advantage of working with this model is that Jeroen is not reliant on AI systems from large tech companies, which might use his data. The drawback is that it takes considerable time to train the model with his own material. Once trained, however, he can use it without limits.

 

To make it clear to the audience when Jeroen is playing and when the AI takes over, a light installation with color codes has been added. This installation is connected to the Signal Interference system and shows the audience which “player” is performing at any moment.

 

Are you a musician, maker, or curious about technology in performance? Check out the open-source code on GitHub, experiment with it, and create your own interactive jam session. Share your results, learn from others, and help develop this ecosystem further.

 

Si Jeroen

 

All show dates can be found here. Video explainers on Signal Interference can be found on Jeroen’s Youtube channel.