Artificial music intelligence, and other parlor tricks

Abstract:
I review some research aiming to build machines that possess
intelligence in a variety of domains, but focusing mostly on music
represented aurally or symbolically. These machines have a variety of
applications to address commercial and public needs, such as music
recommendation and discovery, the management and use of music archives,
music pedagogy and practice, and niche activities like "machine folk
music" (https://themachinefolksession.org/). While many of these systems
appear to have some degree of knowledge about music, scratching below
the surface reveals troubling contradictions that question their
real-world usefulness. This motivates a complete rethinking of
contemporary research values and methods.

Bio:
I received my PhD in 2009 from the Electrical and Computer Engineering
Department at the University of California, Santa Barbara, where I
specialised in digital signal processing and its applications to music
and audio data. I arrived to Europe in 2009 for a Chateaubriand postdoc
at UPMC Paris 6 at the Lutherie, Acoustique et Musique; and then moved
to Denmark in 2010 to become an adjunkt and then lektor at Aalborg
University, Copenhagen in the Department of Architecture, Design and
Media Technology. In late 2014, I moved to London to become a Lecturer
of Digital Media at the School of Electronic Engineering and Computer
Science, Queen Mary University of London. Finally, in July 2018, I
joined the Speech, Music and Hearing division of EECS at KTH as lektor
of Computer Science. My research interests are in the application of
digital signal processing and machine learning applied to music data,
evaluation methodologies, and generative music.