Those who have been reading me here or elsewhere for a long time know that I have often explained that the major technological, economic, ideological, cultural and political changes in our societies are generally preceded by very large-scale disruptions in the practice of music. Because it can explore the realm of possibilities more quickly than other sectors of activity, which are slowed down by their more material components. Thus, we were able to deduce from the appearance of the concert hall, the imminent takeover of the market economy; the birth of the symphony orchestra, the coming of the industrial age; the gramophone, the announcement of the consumer society; atonal music, the coming of the disasters of the twentieth century; and the spread of online music, the emergence of the virtual economy.
Today, a new major change is on the horizon, again faster in music than elsewhere: the arrival of Artificial Intelligence everywhere. A real tsunami.
First, music databases are being built up at high speed, fed by all existing works, and they are beginning to be used to write new works by mixing them together. What’s more, applications (such as Soundful’s AI Music Generator or others) now allow anyone, without prior training, to invent an infinite number of new musical instruments and explore the full range of melodic, harmonic, instrumental, vocal, stylistic, rhythmic and frequency spectrum possibilities.
And, as could have been predicted a very long time ago (cf.Noise), every music consumer will now be able to set themselves up as a composer. Already, today, more than 100,000 new works appear every day and the amount of music available will double every two years. New songs, performed by AI doubles of famous singers, go viral on social networks. For example, a cover of a popular song by an AI copy of the Belgian singer Angèle has very recently had tens of millions of views. Even more, Edith Piaf has been made to sing rap. And we could make Taylor Swift sing Gregorian chant, La Callas soul music, have Joseph Haydn write Gospels, Bach new cello suites; have Ravi Shankar play metal; and even, why not, have a choir of dolphins sing a Bach Mass.
And don’t say it’s just plagiarism: the new is always the bridge between two things that no one had thought of or been able to connect until then.
All this will, more generally, in fact, herald a society where creativity will become infinitely available and where the distinction between producer and consumer and between artist and copyist will disappear. Dizzying.
In this society, the rights of musicians over their works, which will have been used to train these AIs, could be completely wiped out. Already, according to the US Copyright Office, music produced by AI is not protected by copyright; in particular, in the case where AI generates original melodies and lyrics and imitates the voice of a famous singer, there is no violation of the copyright of the artist whose voice has been imitated. Moreover, it is very difficult to find the works that inspired an AI work, unless it is possible to trace all the components of that work in the database, in the same way that one tries, often in vain, to trace the origin of the components of a food product. It is very difficult, but not impossible. If one wanted to.
The great performers, meanwhile, will retain their copyright and their remuneration for their concerts and online listening on the platforms. Meanwhile, the countless new talents made possible by these new technologies will struggle to get the platforms to give them a chance: inundated with new proposals, they will focus their efforts on already-known artists, and will develop with them (this is already the case for some) digital workshops, composing and playing for them.
We might even think like Martin Solveig, that these platforms, anxious to gather the largest possible audience, will have works composed by AI performed by famous people (or their digital doubles) known for something other than music, and who will thus have crossed the barrier to fame. For example, by a famous chef, a great sportsman, a world-famous influencer, a talk-show star.
In this future, visibility alone will define value; fame will become the main asset; and consumers, like concert-goers, will no longer be content to consume a good or a service. They will have to choose between remaining in their anonymous producer/consumer bubble, or belonging to the world of a star, to share in their fame and illusory immortality, or even to rebel against both by taking action.
In a world where countless threats accumulate, and where there is so much to be done to put these dizzying means at the service of creation, knowledge and good, illusion could therefore take precedence over action. That would not be good news. It is not inevitable. For the illusion always stops at the boundaries of reality.
Références
Attali, J. (1977). Noise: The Political Economy of Music. Paris: by Presses Universitaires de France. Récupéré sur https://en.wikipedia.org/wiki/Noise:_The_Political_Economy_of_Music
Image generated by AI (Midjourney)