A.I. Music — It’s About the Song

I’ve been thinking a lot about the argument over A.I. music, and for me it keeps coming back to something very simple: it’s about the song. Does it make you feel something? Does it tell you a story? Does it make you smile, think, cry, dance, remember somebody, or look at the world a little differently? If it does, then the song has done its job.

It's All About the MusicMark Allen Carper

0:00

Recorded music has always relied on technology. Studios have been filled with electronic equipment since long before A.I. arrived: multitrack recording, synthesizers, sampling, compression, reverb, pitch correction, Auto-Tune, digital editing and those enormous mixing desks covered in effects. Nobody listens to a great record and says the emotion somehow became invalid because electricity helped make it. What has changed now is the speed and accessibility of the technology. Something that might once have taken months, a recording studio, musicians, engineers, producers and a lot of money can now begin with one person, an idea and a computer.

And that is where I think the real argument begins. This technology challenges the gatekeepers. For decades, the music industry was controlled by record companies, executives, producers, radio programmers and powerful industry moguls. Think about figures like Sean “P. Diddy” Combs and the enormous influence people at that level could have over careers, contracts, promotion and access to the industry. You once had to get through people like that. Somebody had to decide you were worth recording. Somebody had to finance you. Somebody had to distribute you. Somebody had to decide your music deserved to reach an audience.

Now a person sitting at home can say, “I have something I want to say,” write a song, experiment with it, generate different performances, change the arrangement, throw away the versions they don’t like and keep working until they hear something they love. I know because that is exactly what I’ve been doing. Sometimes I go through generation after generation before I get the result I want, but when I finally hear it, I can genuinely fall in love with the song.

I’ve made hundreds of them. Enough that I built The Felicity Station, an entire radio station around the material I’ve created. I stream it for free. No membership, no subscription and no sign-up. I just wanted somewhere to send these songs out into the world. A lot of them are about love, hope, imagination, people, places and ideas that matter to me. Some are serious, some are fun, some are strange, and some have surprised me with how much emotion they managed to carry.

People sometimes say an A.I. performance cannot contain genuine emotion because the machine itself doesn’t feel anything. But a piano doesn’t feel anything either. Neither does a microphone, a synthesizer or a mixing console. The feeling happens between the idea, the song and the person listening. I have heard songs created through these tools that have stopped me in my tracks because something in the performance hit exactly the emotion I was trying to communicate. Whatever technology helped produce that moment, the feeling I experienced was real.

That doesn’t mean live musicians disappear. I think the opposite may happen. We will always want live performance because there is something completely different about watching another human being stand on a stage and perform in front of you. The connection between performer and audience, the imperfections, the energy in the room and the shared experience of a concert are things recorded music never replaced, and A.I. won’t either.

What A.I. changes is who gets the opportunity to create recorded music in the first place. It means more people can make songs, soundtracks, radio stations, films, games and stories without first asking permission from an industry gatekeeper. Some of it will be terrible. Some will be brilliant. Some songs might reach millions of people and others might reach ten people and mean everything to those ten people. That is how creativity should work.

Pandora’s box is already open anyway. If one company closes a door, another technology will open one. If one platform restricts what creators can do, people will learn another platform. The knowledge has spread, the tools are multiplying and people have discovered that they no longer necessarily need somebody sitting behind a corporate desk deciding whether their idea deserves to exist.

So I’m not particularly interested in fighting over whether something deserves the label “A.I. music,” “real music,” or anything else. Put the song on. Listen to it.

If you love it, you love it.

If you hate it, skip it.

It’s about the song.

The part I keep coming back to is the language being used around high-volume creators. Calling somebody a “bad actor” because they make a lot of music says more about the old model than it does about the creator.

I’ve made hundreds of songs because I enjoy making them. I experiment, rewrite, regenerate, change styles, throw versions away and keep going until I get something I genuinely love. That process has given me enough material to build an entire radio station around it. I stream that music for free because I want people to hear it. No sign-ups, no subscription wall, no attempt to manipulate charts. Just songs.

So if making a lot of music suddenly puts me in the same category as people abusing systems, gaming streaming payouts or running automated spam operations, then the definition is far too broad.

High output does not equal bad intent.

That matters because one of the biggest things A.I. has changed is scale. The old industry model made large-scale production expensive. You needed studio time, musicians, producers, engineers, equipment, distribution and money. That naturally kept most people from producing hundreds of finished songs. Now one person can work at a level that would once have required an entire production operation.

Of course the old model finds that disruptive.

The strange part is treating the disruption itself as suspicious.

For decades, the gatekeepers controlled access. Labels decided who got signed. Executives decided what got funded. Producers and industry figures decided who got access to studios and promotion. Powerful people like P. Diddy existed inside a system where careers often depended on getting through the right doors and dealing with the right people.

A.I. weakens that structure because suddenly I don’t need somebody like that to give me permission to create.

I can write my own lyrics. I can develop an idea. I can generate arrangements. I can try twenty versions if I want to. I can build an entire catalogue. I can put those songs on my own station and let people decide for themselves whether they like them.

That is a massive transfer of creative power.

And I think that is why the phrase “bad actor” bothers me so much. If I were deliberately exploiting a platform, manipulating royalties, impersonating artists or flooding services with automated junk for financial gain, then call that behaviour what it is. But don’t casually group legitimate high-volume creators into the same category simply because we are using the technology more extensively than expected.

The technology has changed what one person can do.

That is the whole point.

We went from a world where making one polished song could take months and thousands of dollars to a world where somebody with an idea can explore dozens of musical possibilities in an afternoon. That means we are going to see more music. A lot more music.

Some of it will be awful. Some of it will be forgettable. Some of it will be wonderful. That has always been true of art.

The difference is that more people finally get to participate.

If the old industry calls that dangerous, excessive or the work of “bad actors,” I think we should ask why.

Maybe the problem isn’t that ordinary people are suddenly making too much music.

Maybe the problem is that the gatekeepers are discovering they no longer control who gets to make it.

Return to The Felicity Station