Music from the Feudal Era and Its Alternatives

Music from the Feudal Era and Its Alternatives

Jacques Attali: the founding president of the European Bank for Reconstruction and Development, former special advisor to French President François Mitterrand, and author of 86 books.

If we want a glimpse of what our AI-driven future may look like, we need only to reflect on what is happening in the world of music. The market for non-scarce products and services has begun through music, as the concept of rewarding intellectual endeavors emerged. Johann Sebastian Bach had to perform concerts in cafés to support his large family. However, with the advent of the Industrial Revolution, mass production, and intense specialization that this revolution brought, much larger markets became accessible.

Now, the digital economy has taken this concept even further. Today, musicians who previously relied on lords and major feudal patrons, then on bourgeois consumers who could afford concert tickets, and finally on record companies that paid them fees, now earn their income from streaming services and other digital platforms.

Now, AI is turning the industry upside down, as generative AI tools can produce music without human composers, drawing from an enormous catalog of existing works for training. For example, the virtual band The Velvet Sundown surpassed the one million mark in listens on Spotify within weeks, and the song “Heart on My Sleeve,” released by an anonymous user on TikTok using AI to create a song for rapper Drake featuring singer The Weeknd, garnered millions of views.

We will also see an AI DJ capable of hosting parties as if they were human, complete with speeches and playlists, in addition to film scores created by AI and voices mimicking artists and their styles. In many cases, virtually anyone can generate music and sounds at a low cost for a wide range of applications.

The reality is astonishing. The International Confederation of Societies of Authors and Composers (CISAC) predicts that the market for AI-generated music and audiovisual content will skyrocket from around €3 billion ($3.5 billion) today to €64 billion by 2028, with AI-generated music potentially accounting for about 20% of streaming platform revenues. CISAC also indicates that the revenues of creators have become endangered; in music, total earnings could fall by roughly 24% by 2028.

To protect copyrighted materials of artists, policymakers in some jurisdictions have begun to take legislative action. The European AI Act requires those who publish and distribute AI-generated materials to be transparent about the sources of that material. In addition, many European projects are exploring watermark-based solutions and blockchain technology to trace source materials and automatically pay small fees. However, it is likely that these protective measures will prove illusory. It will be necessary to reward tomorrow’s artists in different ways. The emergence of a new type of economy means everything must change.

Ultimately, anyone with a computer or mobile phone can create, arrange, mix, master, and produce music videos or adapt their works for video games, interactive advertisements, marketing campaigns, and other uses. One possibility, then, is that generative AI will empower some artists to abandon arrangements with record companies and other traditional intermediaries. In this way, they might try to maintain a personal dialogue with their fans, who can provide customized experiences.

Sensing the impending changes, music distribution platforms are trying to get ahead of the game by allying with established record companies, which themselves face the risk of extinction. For example, Spotify signed an agreement with three major record labels promising to use AI with human artists to their benefit, ensuring transparency, consent, reward, and protection from replicated voices. However, these long-established companies will likely be unable to fulfill their promises, as the reward mechanisms stipulated in these agreements will be largely illusory: they are too small and lack genuine oversight.

Thus, if artists are not cautious, the disruption caused by AI will resemble a change of lord: after the feudal lord, the bourgeois, and powerful record companies will come the algorithmic triumph. Copyright protections will evaporate, and musicians may become mere employees of the algorithm, or even its slaves. The only way artists may escape this fate is by becoming entrepreneurs of their own creations, harnessing the immense potential of AI themselves, and also benefiting from the irreplaceable nature of live musical performance, a reality that is already visible and enduring.

On the other hand, consumers, who may become passive individuals subjected to algorithmic control, can assert themselves. They can become co-composers, determining the form that the work they listen to will take (by choosing the style of music, instruments, and singers), and like the artist, they too become capable of preferring the essential, direct, and live exchange that is integral to musical performance.

The only true freedom, in music as in any other field, is the freedom to create and control the fruits of one’s creativity. AI could enhance this freedom if we act now by focusing on cultivating creativity in schools and beyond. But under the current circumstances, it appears that AI is poised to do the exact opposite.