The argument for AI music usually arrives wearing very attractive clothes.
It promises access. It promises speed. It promises that anyone, regardless of training, money, background, or technical ability, can suddenly make a finished song. That sounds wonderful on the surface. Who would not want more people to feel connected to music?
However, underneath that promise is a far more troubling idea: that music is not a craft to be learned, shared, struggled with, and passed down, it is a consumer experience to be packaged, gamified, monetised, and sold back to us.
That distinction matters.
Suno, one of the most visible generative AI music companies, has positioned its vision of the future around the idea that music should become more like video games. Its CEO, Mikey Shulman, has repeatedly suggested that music is too passive, that it needs to be more interactive, more engaging, more social, more like Fortnite.
On one level, I understand the sales pitch. Video games are active. They are immersive. They generate enormous revenue. Investors understand them. Venture capital understands them. If you can turn music into something users do every day, pay for constantly, and spend hours inside, you have not just built a music tool, you have built a platform.
However, music is already interactive.
It is called playing an instrument.
Music is already multiplayer.
It is called being in a band.
Music is already a deeply engaging experience.
It is called singing with people, playing in a room, writing with a collaborator, arguing over a chord change, finding the right groove, making mistakes, trying again, and finally feeling something click.
The problem AI companies are trying to solve is not that music is passive. The problem they are trying to solve is that craft takes time, and time is difficult to monetise at scale unless you can compress it, automate it, and turn it into a subscription.
That is where the language gets revealing. The AI music pitch is full of words like speed, iteration, engagement, experiences, interaction, and consumption. This is not the language of musicians. It is the language of product design. It is the language of apps, growth curves, investor decks, and customer behaviour.
When you apply that language to music, the danger is obvious. The value of the process disappears.
A song is no longer the result of years of listening, learning, practising, failing, absorbing influences, developing taste, working with others, and discovering a voice. It becomes an output. A result. A finished product generated as quickly as possible.
That might be useful in some contexts. It might even be fun. However, we should not confuse convenience with creativity.
One of the most important points raised in the transcript is that when users were asked what Suno allowed them to do that traditional instruments or DAWs did not, the most common answers were not musical. They were practical. It saves time. It saves money. It works like a collaborator.
That says everything.
The selling points are not harmony, melody, groove, feel, emotion, taste, or expression. They are speed, cost reduction, and the replacement of other people.
That is a profoundly lonely vision of music.
Real collaboration is not just having something give you options. A collaborator challenges you. A collaborator brings their own taste, history, limitations, brilliance, and stubbornness into the room. A collaborator can tell you when something is not good enough. A collaborator can frustrate you, surprise you, and make you better.
AI does not do that. It flatters. It produces. It gives you something back. You accept, reject, regenerate, and continue. There is no negotiation. There is no accountability. There is no human being on the other side whose instincts you have to respect.
That is not collaboration. That is ordering.
And ordering food does not make you a chef.
The same is true of the argument around taste. We are increasingly told that skill matters less now, because taste is what really counts. That sounds sophisticated, especially when people bring up someone like Rick Rubin, who is famous for saying he knows what he likes and what he does not like.
However, that argument badly misunderstands what taste is.
Taste is not something that floats above craft. Taste is developed through craft. You hear differently after years of recording. You hear differently after trying to play like your heroes and failing. You hear differently after tuning vocals, editing drums, choosing microphones, balancing a mix, writing a chorus that does not work, rewriting it, then finally finding the line that does.
Taste is not merely selection. Taste is perception shaped by experience.
That is why role models matter. Musicians grow by looking up to other musicians. We hear Jimi Hendrix, Jaco Pastorius, Stevie Wonder, Queen, The Beatles, Prince, Joni Mitchell, or whoever opened the door for us, and we think, “How did they do that?” That question is the beginning of a life in music.
AI prompting often removes that question. There is no hand to watch, no breath to hear, no room to imagine, no human decision to study. There is only output.
That is why the issue of deskilling is so serious. If musicians start relying on AI systems to make creative decisions for them, the danger is not simply that the work changes. The danger is that the musician changes. The muscles of decision making weaken. The ear becomes passive. The instinct becomes outsourced.
In music, the slow part is often the meaningful part. Learning patience is part of learning how to make anything worthwhile. Sitting with an idea, wrestling with it, getting annoyed with it, abandoning it, returning to it, and finally understanding what it wants to become, that is not wasted time. That is the work.
AI culture often treats friction as a problem. However, in art, friction is frequently where the identity is formed.
Of course, there may be ethical and interesting uses for generative AI in music. As a memorisation tool, it could help people retain information through melody. In certain therapeutic contexts, it may provide comfort or connection. For someone who has always felt ashamed to make music, AI might even act as a doorway into creativity.
Those possibilities should not be dismissed.
However, they also should not distract us from the larger business model. This is not simply about helping people sing again. This is about building platforms that own the tools, control the experience, shape the behaviour, and monetise the user’s desire to feel creative.
That is not democratisation in any meaningful sense. If the people do not own the means of production, if the platform can disappear, change the rules, raise the price, restrict access, or decide what kind of music gets promoted, then we have not democratised music. We have rented creativity from a tech company.
Real democratisation would mean funding music education. It would mean giving children access to instruments, teachers, rehearsal spaces, choirs, studios, and communities. It would mean rebuilding the casual, communal relationship with music that so many people lose as they get older.
The tragedy is that the social problem AI music identifies is real. Many people feel locked out of music. Many people believe they are not talented enough. Many people were shamed out of singing, playing, or creating. That is heartbreaking.
However, the answer should not be to sell them a machine that does the work for them.
The answer should be to remind them that they are allowed to make music badly at first. They are allowed to learn. They are allowed to be beginners. They are allowed to develop a voice. They are allowed to discover that the joy was never only in the finished song.
The joy was in becoming the kind of person who could make it.
That is what AI music, at its worst, threatens to remove. Not just jobs. Not just income. Not just copyright. Those are huge issues, of course. However, the deeper threat is philosophical. It asks us to accept a world where the appearance of creativity is enough, where the result matters more than the person, where convenience replaces craft, and where human musical experience is reduced to a meaningful consumption experience.
Music deserves better than that.
Musicians deserve better than that.
And listeners deserve better than that.
The future of music should not be built by people who think the problem with music is that it is not enough like a video game. It should be built by people who understand that music is one of the oldest, deepest, most human things we do.
Not because it is efficient.
Because it connects us.
