THE ARGUMENT
Real Musician. Real Producer. Real Music.
The Music Is the Point.
The Method Is Part of the Conversation.
Let’s get the disclosure out of the way.
I use generative AI in the production of Sheridan Rhodes music.
That does not mean I type “smooth jazz” into a box, download whatever comes out, and call it mine.
The music often starts with me playing.
Bass. Guitar. Rhythm parts. Musical ideas recorded and developed the same way I have approached music for decades: I hear something, play with it, and try to figure out where it wants to go.
Then I bring generative AI into the process.
I use it to develop ideas, explore instrumentation, build arrangements, test directions, and take the original material into places I may not have been able to reach alone.
Sometimes the AI contribution becomes a major part of the finished track. Sometimes it supports something I already built. Sometimes it gives me ten answers to a question and I reject all ten.
This is collaboration with a generative system.
The starting idea is mine.
The live parts are mine.
The direction is mine.
The decisions are mine.
And the finished record exists because I stayed involved from beginning to end.
I’m telling you plainly because disclosure is not an apology. It is information.
Now we can have the more interesting conversation.
The Machine Is Not the Musician
One of the strangest things about the AI music debate is how quickly people assign the machine a personality.
AI wrote the song. AI made the music. AI decided. AI replaced the artist.
No.
Generative AI produces output in response to material and direction.
In my case, that process may begin with live bass, guitar, rhythm parts, a lyric, or a musical idea I have already started developing. AI becomes part of the production process rather than the entire source of the music.
It can expand an arrangement. Suggest another direction. Introduce instrumentation. Reinterpret material. Generate performances and variations at a speed that would have seemed ridiculous when I started recording music in the late 1990s.
What it cannot do is give me a reason to make something.
It doesn’t know why a certain Rhodes sound reminds me of Chicago at night. It doesn’t know what I lost. It doesn’t know why an old lyric has been sitting in my head since 1996. It doesn’t know why I played the bass part that way in the first place.
And it doesn’t get dissatisfied when a perfectly competent result completely misses what I was trying to hear.
I do.
That’s where the work starts.
The machine can contribute.
I still have to decide what deserves to survive.
AI Didn’t Give Me an Ear
I’ve been around music production since about 1980.
I played. I wrote lyrics. I worked with recording equipment. I learned to listen to arrangements.
I learned that adding another part is often easier than admitting the song needs one removed. I learned that musicians can play something beautifully and still play the wrong thing for the record. I learned that a technically imperfect moment can sometimes carry more life than the polished replacement.
I still play.
A Sheridan Rhodes track may begin with my bass, my guitar, or rhythm parts I recorded before generative AI ever enters the room.
The technology becomes a collaborator in developing the music.
It did not teach me how to hear it.
AI didn’t give me an ear. It gave my ear a new set of tools.
That’s the distinction I keep coming back to.
“But AI Is Different”
Yes.
It is.
That’s another argument I have no interest in dodging.
Generative AI is not just a new guitar pedal. It isn’t a drum machine with more presets.
The system can actively contribute musical material in response to direction and existing work. It can develop a live rhythm part into a broader arrangement, explore instrumentation around a bass line, or take an unfinished musical idea into directions I had not considered.
That dramatically changes the distance between an idea and the possible finished record.
It also creates legitimate questions about authorship. It creates legitimate questions about copyright, training data, compensation, displacement, and the economic future of musicians.
Those conversations need to happen.
What I reject is the lazy version of the debate where somebody hears “AI was involved” and immediately considers the case closed.
The better question is what the person actually did.
Did they play?
Did they bring an idea?
Did they write anything?
Did they establish direction?
Did they revise?
Did they arrange?
Did they edit?
Did they reject weak material?
Did they develop a recognizable sound?
Did they make choices based on memory, experience, humor, grief, restraint, or taste?
Are they present in their own work?
Those are the questions worth asking.
Why Sheridan Rhodes Exists
Sheridan Rhodes is a music project.
The music comes first.
But this site isn’t only here because I may eventually sell CDs. And it isn’t a digital trophy case I built so I can stare lovingly at my own album covers.
Sheridan Rhodes is also a platform.
A soapbox.
A podium, when necessary.
I found my way back to making music through a combination of the instruments I already knew and generative technology that gave me new ways to develop what I played.
That makes this conversation personal for me.
These tools did not replace my musical life.
They became collaborators in rebuilding it.
So yes, I am going to advocate for serious AI-assisted, AI-augmented, and AI-collaborative creative work.
I am going to challenge lazy “AI slop” arguments when the label is being used as a substitute for listening. I’m also going to criticize lazy AI creators when they confuse volume with authorship.
I will talk about tools. I will talk about process. I will talk about good AI music, bad AI music, disclosure, gatekeeping, authorship, and whatever strange argument the music industry wanders into next.
I’m not neutral on the integration of generative AI into music production.
I’m doing it.
I’m playing the parts.
I’m developing the ideas.
I’m working with the technology.
And I’m listening to the results with the same ears I’ve been using for decades.
That’s the conversation I’m interested in having here.
Well, actually on Substack.
New tracks, and more on the case for AI in music, live on the blog.