For whom the algorithm tolls? 

With the recent developments of AI in music, the automation of analysis, synthesis and composition seems more and more achievable. Once again, a new technology is reviving the utopia of the exclusion of the interpretive-human element from the creation process. Computer science research today is driven by the idea that AI systems can satisfy a need on the part of the music industry – especially the applied music industry (television, advertising, sound design, cinema and popular music) – for alternative and efficient systems for the rapid creation of tunes that can be immediately used in production processes.

We could ask ourselves: what about musical creativity? What about the author’s intention? Can we really believe the narrative of human-machine interaction whereby, now ousted from the system, composers should limit themselves to asking the system what to produce, and accept the frightening corollary that the ‘best’ authors are those who can ‘best communicate their intentions to the machine’?
The Kierkegaardian angst that leads many composers and musicologists to reject the new developments of AI a priori, coupled with a utilitarian and capitalist vision of technology, is increasingly leaving the humanistic world on the sidelines from what will be, if one can venture to predict, practices of music creation that will be increasingly used in many areas of artistic production. But a look at music in historical perspective shows us a cyclical return of fears and enthusiasms towards (semi-)automatic systems of composition (from the isorhythmic motet to integral serialism) and the irruption of technology in music production (electronic music); lapses of high tension that, if it is true that they led some artists to succumb, they also stimulated others to create new and very fine artworks. It is then a matter of intuiting the folds of these new technologies where imagination can draw new stimuli, and new ideas can flourish. Creation of datasets, definition of models for feature extraction, original manipulation of features: the possibilities that open up are manifold. At that point, like so many 20th century authors did (Krenek, Risset, Romitelli, to name but a few), composers and musicologists can return into the technological system, be its creators, its watchmakers. Thus, creativity will no longer be subjugated, nor relegated to the myth of human-machine interaction, but it will finally be able to perform the most recent of its innumerable peripluses and to reborn once again renewed.