Select Page

music.producer // ai.researcher

i produce music and do research in artificial intelligence


Anna’s Brains

First Dutch hip hop song written by AI

I appeared on VPRO’s Anna’s Brains on NPO3 to find out if we can make the first Dutch hip hop song written by AI for artist Daan Boom. The track was written by AI, and produced by me. You can learn more about the project here.

Klaas Kan Alles

Can we make a hit song with AI?

Klaas invited me on his show to embark on the challenge: can we write a hit song with AI? Without anyone knowing this track was written by AI or performed by Klaas, we submitted the track to The Netherlands’ most popular “music charts gatekeeper”, ‘Maak het of Kraak Het’. You can find out if our song made it on the charts here.

Tears on the Dancefloor:

A tearjerker written with Artificial Intelligence

In collaboration with Utrecht and Pure Ellende, I designed the music and lyrics composition software that was used to create the first Dutch “smartlap” song. You can read more about the project at

tv // radio



Contrastive Learning of Musical Representations

Janne Spijkervet and John Ashley Burgoyne, Proceedings of ISMIR 2021

link to paper


While supervised learning has enabled great advances in many areas of music, labeled music datasets remain especially hard, expensive and time-consuming to create. In this work, we introduce SimCLR to the music domain and contribute a large chain of audio data augmentations, to form a simple framework for self-supervised learning of raw waveforms of music: CLMR. This approach requires no manual labeling and no preprocessing of music to learn useful representations. We evaluate CLMR in the downstream task of music classification on the MagnaTagATune and Million Song datasets. A linear classifier fine-tuned on representations from a pre-trained CLMR model achieves an average precision of 35.4% on the MagnaTagATune dataset, superseding fully supervised models that currently achieve a score of 34.9%. Moreover, we show that CLMR’s representations are transferable using out-of-domain datasets, indicating that they capture important musical knowledge. Lastly, we show that self-supervised pre-training allows us to learn efficiently on smaller labeled datasets: we still achieve a score of 33.1% despite using only 259 labeled songs during fine-tuning. To foster reproducibility and future research on self-supervised learning in music, we publicly release the pre-trained models and the source code of all experiments of this paper on GitHub.





That’s The Spirit!

Written for the Metropole Orchestra


Marimba, Vibrophone & String Quartet


there is a vitality

a life force

a quickening that is translated through you into action

and because there is only one of you in all time

this expression is unique

but no artist is pleased

there is no satisfaction whatever at any time

only a queer divine dissatisfaction

a blessed unrest that keeps us marching

and makes us more alive

than the others

– Martha Graham