By Ephraim Agbo
For generations, discovering a new musician meant discovering a person. There was usually a face behind the voice, a story behind the lyrics and a human being making choices about what to say, how to say it and how the music should sound. That relationship between artist and audience has been one of the foundations of popular music. But artificial intelligence is beginning to complicate that relationship in ways the industry is only starting to understand.
Today, a song can sound deeply personal without necessarily having a personal origin. A voice can sound vulnerable without belonging to a human singer. An artist can have a photograph, a name, a catalogue of songs and even an apparently coherent public identity without actually existing as a person. Artificial intelligence is no longer simply assisting musicians in the production of music. In some cases, it is being used to create the music, the voice and the artist itself.
That development raises a question that goes beyond whether AI can make a convincing song. It asks us to reconsider what we mean when we call someone an artist, what audiences have a right to know and how the economics of music may change when songs can be produced at a scale no human musician could possibly match.
The Economics of Endless Music
The most disruptive feature of generative AI may not be the quality of the music it produces, but the speed and scale at which it can produce it.
A human musician may spend weeks or months developing an album. There are lyrics to write, melodies to refine, instruments to record, vocals to perform, arrangements to reconsider and mixes to perfect. Behind a finished record is often a long chain of creative decisions involving songwriters, producers, musicians, engineers and performers.
Generative AI can compress parts of that process dramatically. A user can provide a description of a desired song—its genre, mood, subject, instrumentation or vocal character—and an AI system can produce a musical result within minutes, sometimes seconds. The technology continues to improve, making the boundary between a professionally produced recording and a synthetic one increasingly difficult for casual listeners to identify.
That creates an unusual economic problem. Human creativity operates at the speed of human beings; artificial intelligence operates at the speed of computation. A musician can write only so many songs in a day. An automated system can generate an enormous catalogue without needing sleep, studio time or the physical limitations that govern human production.
The consequence is not necessarily that listeners will abandon human musicians. The more immediate concern is that the amount of music entering the digital ecosystem could become so enormous that attention becomes even more difficult to secure. The human musician is no longer competing only with other musicians. Increasingly, they may be competing with systems capable of producing thousands of songs at a fraction of the cost.
This changes the fundamental economics of recorded music. When music becomes almost infinitely abundant, the scarce commodity is no longer the song. It is the listener's attention.
AI Is Not the Problem by Itself
It would be misleading to suggest that artificial intelligence is inherently harmful to music. Technology has always transformed the creative process.
The synthesiser changed the possibilities of sound. Digital recording transformed production. Computers made editing and composition more accessible. Software allowed musicians without expensive studios to produce professional-quality recordings from their homes.
AI can similarly become another instrument in the creative process.
A songwriter may use it to explore an arrangement. A producer may use it to experiment with sounds or generate musical ideas. An independent artist who cannot afford a large production team may use AI-assisted tools to develop a song that would otherwise remain an unfinished idea. There is nothing inherently dishonest about that.
The more difficult question concerns transparency.
There is an important difference between an artist saying that artificial intelligence helped create a song and an entirely artificial artist being presented to listeners as though they were a real human musician.
The first is technological assistance. The second involves the manufacture of identity.
That distinction is likely to become increasingly important as AI-generated artists become more sophisticated.
When the Artist Itself Is Manufactured
Imagine discovering an unfamiliar musician on a streaming platform. The voice is compelling, the songs are memorable and the artist appears to have a distinctive visual identity. You follow the account, listen to the music repeatedly and perhaps begin to identify with the story supposedly behind the artist.
Later, you discover that the photograph was generated, the voice was synthetic and the biography was fictional. There was no singer living the life described in the artist's profile. There was no childhood behind the story, no studio session in which the performance took place and no human performer whose experiences formed the foundation of the identity being marketed.
The music may still be enjoyable. That is not the central issue.
The deeper issue is that the audience was not simply consuming music. It was consuming an identity, and the identity turned out to be something manufactured.
This is where transparency becomes important. Listeners are entitled to make different choices for different reasons. Some may care deeply about supporting human musicians. Others may be interested only in whether a song sounds good. Neither position is inherently wrong. What matters is that listeners have enough information to make that choice knowingly.
But Authenticity Is More Complicated Than That
There is also a temptation to argue that AI-generated music is inherently less authentic because a machine cannot experience love, grief, poverty, heartbreak or any of the other emotions that inspire music.
That argument, however, does not fully withstand scrutiny.
Human musicians have never needed to personally experience every story they sing about. Songwriters routinely write songs for other performers. A singer may deliver lyrics written entirely by somebody else. A musician can write about war without having fought in one, or about migration without having personally crossed a border. Actors portray experiences they have never lived, and composers can create music that expresses grief without having experienced the specific tragedy that inspired the composition.
Music has never required the performer to have personally lived every word.
The more meaningful questions are therefore about authorship, intention and representation. Who created the work? Who made the creative decisions? Who owns it? Who is performing it? And, crucially, is the audience being given an accurate understanding of the process?
A human songwriter writing a heartbreak song for another singer is not deceptive simply because the singer did not write it or personally experience the exact situation described in the lyrics. Likewise, a musician using AI as one component of the production process is not automatically compromising the authenticity of the work.
The situation becomes fundamentally different when an entire artistic identity is fabricated and presented as human when it is not.
The issue is not that a machine has never experienced heartbreak. It is that a machine can now help create the appearance of a person who has experienced heartbreak when no such person exists.
Streaming Platforms Face a New Challenge
The emergence of synthetic artists is forcing streaming companies to confront questions they were not originally designed to answer.
For years, platforms have largely treated the artist name, song title, album and audio recording as the essential pieces of information surrounding a piece of music. Artificial intelligence complicates that model because the process through which the music was created can now be commercially and culturally significant.
Spotify, for example, has announced an "AI Persona" label for certain entirely AI-generated artist identities. The move reflects an emerging industry effort to give listeners more information about synthetic performers and distinguish them from conventional human artists.
Other streaming companies are also experimenting with ways of identifying AI-generated music. The broader direction is clear: the technology behind a recording is increasingly becoming part of the information that platforms may need to disclose.
But labelling will not be simple.
There is a significant difference between a completely AI-generated song and a human recording that contains one AI-generated element. A musician might use AI to create a backing track while writing and performing the vocals themselves. Another might use an AI system for harmonies, while everything else is performed by humans. Someone else might generate almost an entire song and then make substantial human changes.
Where should the boundary be drawn?
A simple "AI" or "human" label may eventually prove inadequate. The industry may need a more detailed system that explains not merely whether AI was used, but how extensively it was used and what role it played in the creative process.
The Algorithm Makes the Problem Bigger
There is another reason this matters: streaming platforms do not merely store music. Their recommendation systems increasingly influence what people discover.
A song can exist on a platform without ever reaching a significant audience. But once an algorithm begins recommending it through personalised playlists, search results or other discovery systems, its potential reach changes dramatically.
This is where the sheer productivity of generative AI becomes economically significant.
A human artist might spend months producing an album. An AI-driven operation could potentially generate hundreds of tracks during the same period. If recommendation systems reward engagement, frequency or other measurable signals, large-scale synthetic production could create new opportunities for manipulation.
The concern is not simply that AI music might become popular. It is that the economics of the streaming environment could reward whoever can produce the greatest volume of content at the lowest cost.
That would put independent human musicians in an uncomfortable position. Their greatest competitive disadvantage would not necessarily be a lack of talent, but the fact that they are human.
Why Nigeria Should Pay Attention
The issue has particular significance for Nigeria.
The Nigerian music industry has become one of the country's most visible cultural exports. Afrobeats has created international opportunities for singers, songwriters, producers, engineers, managers, dancers and a vast network of professionals whose livelihoods depend directly or indirectly on music.
Artificial intelligence could create new opportunities within that ecosystem, particularly for young and independent creators who lack access to expensive production facilities. But it could also create serious challenges involving copyright, ownership, voice imitation and compensation.
Nigeria's existing copyright framework was developed before the current generative AI explosion. That leaves difficult questions that will increasingly demand clearer answers. Who owns a song generated with artificial intelligence? Can an AI system legally be trained on copyrighted recordings? What happens when a system generates a voice that sounds like an established Nigerian artist? Can a musician prevent the commercial use of an artificial version of their voice? And who receives the royalties when an AI-generated recording becomes successful?
These questions become even more important when the technology moves beyond imitation and begins producing entirely new musical identities.
Imagine an artificial artist whose voice resembles a famous Nigerian performer, whose songs are marketed to Nigerian listeners and whose music begins accumulating millions of streams. The artist never recorded the songs. The voice is synthetic. The identity may be fictional.
Who, then, is entitled to the money?
The answer cannot simply be left to technology companies and streaming algorithms. It will require musicians, lawyers, regulators, platforms and audiences to decide what rules should govern the emerging market.
The Voice May Become the New Battleground
The human voice is one of the most recognisable aspects of an artist's identity. Fans can identify favourite singers within seconds, often without seeing their faces.
That makes voice cloning particularly powerful.
An AI system capable of generating a convincing imitation of a singer can potentially create performances that the artist never recorded. The technology therefore raises questions that extend beyond copyright. It touches personality rights, consent, reputation and commercial identity.
The distinction between inspiration, imitation and impersonation becomes increasingly difficult when technology can reproduce not merely the general characteristics of a musical style, but aspects of an identifiable person's voice.
For musicians, this could make the voice itself one of their most valuable intellectual and commercial assets.
The Real Battle May Be About Trust
For decades, the music industry has sold more than sound. It has sold identity.
The mysterious singer, the rebellious performer, the songwriter with a complicated past, the street poet, the romantic crooner and the superstar with a recognisable face and unmistakable voice are all part of the way music has been marketed.
Artificial intelligence can now manufacture convincing versions of those identities.
That is why the AI music debate is ultimately larger than technology. It concerns trust, labour, copyright, money and the relationship between audiences and the people—or systems—creating the music they consume.
The question is not whether machines should be allowed to make music. They already can.
The question is what rules should govern an industry in which a song can be created without a conventional musician, a voice can exist without a singer and an artist can become famous without ever having existed.
What Happens to Human Creativity?
It would be a mistake to assume that the rise of AI means the end of human creativity.
Human beings have always adapted to new technologies. Some musicians will reject AI entirely. Others will incorporate it into their creative processes. Some songs will be entirely human. Others will be collaborations between humans and machines. Some will be almost completely synthetic.
The boundaries will become increasingly blurred.
And perhaps that is why the conversation should move away from the simplistic idea of "human versus AI."
The more useful question is whether audiences, artists and platforms can develop a system in which technology expands creativity without making deception the business model.
A human artist does not have to write every lyric personally to make meaningful music. A singer does not have to experience every emotion described in a song. And using AI does not automatically make a piece of music dishonest.
What matters is the integrity of the relationship between creator, platform and audience.
Listeners should know what they are being offered. Artists should know when their voices or creative identities are being used. Creators should have a meaningful opportunity to receive compensation when their work contributes to the development of commercial systems. And platforms should not allow industrial-scale synthetic production to quietly overwhelm human creators simply because machines can produce more content at lower cost.
A New Definition of the Artist
The music industry may eventually have to reconsider what the word "artist" means.
Is an artist the person who writes the lyrics?
The person who performs the vocals?
The producer who constructs the recording?
The person who writes the prompt?
The company that owns the AI system?
The machine that generates the final performance?
Or some combination of all of them?
There is no easy answer yet.
What is certain is that the old assumptions are becoming harder to maintain.
For most of modern music history, the technology could change how a person made a song without fundamentally changing the fact that there was a person making it.
Generative AI breaks that assumption.
It introduces the possibility that the music, the performance and even the artist's identity can all be manufactured.
That does not necessarily make the resulting music worthless. Nor does it mean human music will automatically be better.
But it does mean that audiences will increasingly need to ask questions they rarely had to ask before.
Not simply whether they like a song, but who made it.
Not simply who appears on the artist profile, but whether that person actually exists.
Not simply whether AI was involved, but how it was involved.
And ultimately, whether the person—or system—being presented to them is what the platform says it is.
The future of music may therefore not be a battle between humans and machines.
It may be a negotiation between creativity and scale, technology and ownership, convenience and transparency, and ultimately between what sounds human and what is genuinely human.
As artificial intelligence becomes better at making music, the value of human creativity may not disappear.
It may simply become harder to see—and therefore more important to protect.
No comments:
Post a Comment