Collegium Musicæ Seminar "Analysis l Creation" - Automatic Learning Applied to Music

If we apply the famous Amara’s law (“We tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run.”) to the current wave of technical issues surrounding automatic learning, doubtless, the first premise is the most realistic. However, it is also true that, despite the indisputable results of techniques such as deep learning, there is an unwillingness to accept these newcomers in our artistic and scientific worlds: the arguments vary from the fear of replacing man by a machine to the traditional “nobody knows why it works!”.

During this seminar, we will study machine learning applications for music. But, rather than pass judgement or simply find out if it works, we will demonstrate how the intrusion of these techniques could make it possible to understand musical activity and its future. Several ethereal notions, like artistic creation for example, could become less elusive through the arrival of artificial intelligence in the process of musical creation