Sunday Times 2
Serf music and its alternatives in AI era
View(s):By Jacques Attali, PROJECT SYNDICATE, EXCLUSIVE TO THE SUNDAY TIMES IN SRI LANKA
PARIS – For a glimpse of what our AI-driven future could be, consider what is happening in the music business. It was through music that the market for non-rare products and services began, where remuneration for the works of the mind was first imagined. J.S. Bach had to hold coffeehouse concerts to support his vast family, but with the Industrial Revolution, mass production, and the extreme division of labour that it brought, much larger markets became available.
The digital economy took this further. Musicians who once relied on the lords who commissioned them, then on the bourgeois consumers who bought concert tickets, and finally on the record companies that paid them royalties, today are remunerated by streaming services and other online platforms.
Now, AI is turning the industry on its head. Generative AI tools can produce music without human composers, using the immense catalogue of existing works to train themselves. The virtual band The Velvet Sundown passed the one-million mark in streams on Spotify in a matter of weeks, and “Heart on My Sleeve”, posted to TikTok by an anonymous user who “used AI to make a drake song ft. the weeknd,” has racked up millions of views.
One can also find artificial DJs capable of hosting a party like a human, complete with speeches and playlists, as well as AI-generated film soundtracks and voiceovers that imitate artists’ voices and styles. In each case, pretty much anyone can generate low-cost music and audio for use across a broad range of applications.
The evolution is dizzying. The International Confederation of Societies of Authors and Composers (CISAC) expects the market for music and audiovisual content created by generative AI to skyrocket, from around €3 billion ($3.5 billion) currently to €64 billion by 2028, with generative AI music possibly accounting for approximately 20% of streaming platforms’ revenues. CISAC also notes that creators’ revenues are at risk; for music, the total could fall by about 24% by 2028.
To protect artists’ copyrighted material, policymakers in some jurisdictions are beginning to take legis-lative action. The European AI Act requires those who publish and distribute AI-generated material to be transparent about its sources. There are also several European projects exploring watermarking and blockchain-based solutions to identify source material and automatically pay out micro-royalties. But such protections will likely prove illusory. The artists of tomorrow will have to be remunerated in other ways. The arrival of a new kind of economy means everything must change.
After all, anyone with a computer or mobile phone can create, arrange, mix, master, and produce a music video or adapt their own works for video games, interactive advertising, marketing campaigns, and other uses. One possibility, then, is that generative AI will further enable some artists to forego arrangements with record companies and other traditional intermediaries. In doing so, they may try to maintain a personalised dialogue with their fans, whom they can offer customised experiences.
Sensing the changes that are coming, music distribution platforms are trying to get ahead of the game by allying themselves with legacy record companies, which are themselves in grave danger of extinction. For example, Spotify has signed an agreement with three major record companies promising to use AI with and for human artists, thus guaranteeing them transparency, consent, remuneration, and protections against cloned voices. But these legacy players will be unable to keep their promise, because the remuneration mechanisms provided for in these agreements will be largely illusory: too small and without real control.
Thus, if artists are not careful, the upheaval introduced by AI will amount to a change of master: after the feudal lord, the bourgeois, and the all-powerful record company will come the triumph of the algorithm. Copyright protections will evaporate, and musicians will become mere employees of the algorithm, if not its slaves. The only way that artists can escape this fate is by becoming entrepreneurs of their own creations, harnessing AI’s formidable potential themselves, and also capitalising on the irreplaceability, already visible and lasting, of the in-person concert performance.
Meanwhile, consumers, who could become passive subjects of algorithmic control, could assert them-selves. They could become co-composers, determining the form to be given to the work they listen to (by choosing the music style, the instruments, and the singers) and, like the artist, privileging the actu-al, direct, living, irreplaceable exchange of the concert performance.
The only true freedom, in music as elsewhere, is to create and control the fruit of one’s creation. AI could amplify this freedom if we act now by focusing on the development of creativity at school and elsewhere. But as matters stand, it seems well on its way to doing the opposite.
(Jacques Attali, Founding President of the European Bank for Reconstruction and Development, is a former special adviser to French President François Mitterrand and the author of 86 books.)
Copyright: Project Syndicate, 2025. www.project-syndicate.org