The Machine Didn’t Plant the Seed

There is something bothering me about the conversation we’re having around artificial intelligence.

We’re giving the machine too much credit.

Sometimes we’re giving it credit for what’s good. Sometimes we’re blaming it for what’s bad. Either way, we’re forgetting about the person sitting in front of it.

I’ve been writing songs for years. These days, I use artificial intelligence as part of that process. And I’ve discovered something along the way:

The machine didn’t plant the seed. I did.

I Call It “Mama Jammin’”

I have my own expression for the way ideas come to me. I call it “mama jammin’.”

I may hear somebody say a phrase. Remember something that happened 50 years ago. See two people talking, think about someone I’ve lost, or simply put two words together that suddenly mean something to me.

Then I start talking.

And I mean that literally.

I don’t type much anymore. I use the voice-to-text feature of AI and simply talk to it. I talk through an idea the same way I might have talked it through with another person sitting across the table from me.

Sometimes I talk for quite a while. I describe what I’m seeing in my mind, throw out phrases, change direction in the middle of a thought, remember something else and follow that instead.

I don’t necessarily know where I’m going when I begin.

I’m thinking out loud.

That’s what I call mama jammin’.

And somewhere in all that talking, I’ll recognize something worth pursuing.

That’s the seed.

AI didn’t wake up one morning and decide to write a song about Butch Cassidy and the Sundance Kid. It didn’t know about a business partner I once had or what that relationship meant to me.

I brought that to the conversation.

Then AI may take my thought somewhere I hadn’t considered. Sometimes I think, “That’s exactly what I meant.” Other times, “No, that’s not it at all.”

So I change a word. Rewrite a line. Throw out a verse. Change the perspective. Ask for another approach.

That’s creation to me.

It’s a conversation between an idea and a tool.

Owning a Guitar Doesn’t Make You a Guitar Player

I’m not a great guitar player. That’s probably one reason I admire someone who really knows how to play one. I appreciate the years it took to learn the instrument.

And that brings me back to AI.

You can buy a beautiful guitar tomorrow. That doesn’t make you a guitarist.

You can sit down at a Steinway piano. That doesn’t make you a pianist.

And paying for an AI subscription doesn’t make you a songwriter.

Someone can type, “My girlfriend left me. Write me a sad country song.”

The machine will probably produce something. But there’s a difference between generating something and developing something. That’s the distinction I think we’re losing.

The Prompt Became Part of My Instrument

When I first started generating music with AI, my prompts were pretty basic. And I got some good songs. But over time I discovered that the quality of what I received depended enormously on the quality of what I asked for.

So I started learning.

Instead of simply asking for a blues song, I began thinking about tempo, key, chord movement, instrumentation, introductions, solos, melodic structure, spaces between instruments and where things should happen within the song.

Those decisions didn’t come with my subscription.

I had to learn them.

The prompt became part of the instrument.

Here is an example of what one of my prompts looks like today:

• Style / Feel: Slow reflective atmospheric ballad with spacious arrangement and emotional restraint. Late-night mood with cinematic loneliness and quiet warmth underneath.
• Tempo: Slow pulse around 66 BPM with natural breathing space between phrases.
• Key / Tonal Center: G minor with subtle Dorian color to create a darker emotional tone while still allowing moments of lift.
• Intro / Riff: Sparse fingerpicked guitar motif beginning on the minor third, moving downward first before resolving upward in a delayed cadence. Use open-string voicings and suspended tones so the riff feels drifting and unresolved.
• Verse Harmony:

Gm9 → Ebmaj7 → Bbadd9 → Fsus2

Wide, airy voicings with ringing sustain and gentle harmonic movement.

• Chorus Harmony:

Eb → Bb → F → Gm

A slight emotional lift with broader chord spacing while remaining restrained and reflective.

• Bridge:

Cm7 → Eb → F → Gm

A gradual emotional rise adding tension before returning to the quieter mood.

• Instruments: Fingerpicked acoustic guitar foundation, distant electric guitar swells with long delay and reverb, soft upright piano accents, warm analog pads, brushed percussion with light kick drum and subtle ambient textures.
• Melody: Gentle conversational phrasing with long sustained notes followed by breathing space. Melody rises slightly in the chorus but remains intimate and controlled.
• Lyrics / Theme: Quiet emotional realization when seeing someone who quietly changes your world. Reflection, vulnerability and the sense that something profound is happening beneath the surface.
• Vocal: Close intimate female vocal with slight fragility. Natural breathing in the phrasing, understated emotion, no belting or dramatic performance.
• Production: Avoid repeating melodic contours from previous generations. Emphasize a new melodic arc and subtle phrasing variations.

That’s quite a distance from, “Write me a sad song.”

I’m not suggesting everyone needs to understand what Gm9 or Fsus2 means before using AI. My point is that as I learned more about the language of music, I gained more control over what I was asking the machine to create.

The better I became at describing what I heard in my head, the better chance the machine had of helping me get there.

One Song Taught Me a Lot

One of my songs is called “I’m Done.”

I didn’t simply ask AI to make something that sounded good.

I started by studying the musical construction of a successful song I admired. I wanted to understand what was happening underneath the recording.

Then I experimented.

Change this. Move that. Keep this idea but approach it differently. Try another rhythm. Change the arrangement.

Eventually, the original reference disappeared into the process and something new emerged.

We combined that musical direction with my lyrics and produced “I’m Done.”

That song has since become one of the most requested Blue Smoke Band songs on a Las Vegas radio station.

Was AI involved?

Absolutely.

But AI didn’t have the original idea. It didn’t decide what I wanted to communicate. It didn’t listen to the result and say, “Almost—but not quite.”

I did.

Mass Production Isn’t the Same as Creativity

There is another side of AI-generated music that bothers me.

Because generating music has become inexpensive and fast, people can produce enormous quantities of it.

Generate. Download. Upload. Repeat.

Eventually, the world gets flooded with music nobody spent much time thinking about, and that can become the public’s impression of AI music.

But that isn’t a failure of the instrument. It’s a decision made by the person holding it.

Someone can bang randomly on a piano all afternoon. That doesn’t prove the piano is incapable of producing beautiful music.

Owning an instrument and mastering one are two entirely different things.

Maybe We’re Asking the Wrong Question

People frequently ask whether something was created by AI. Maybe that’s becoming the wrong question. Perhaps we should be asking:

• What did the human being bring to it?
• Where did the original idea come from?
• What decisions were made?
• What was rejected?
• What was changed?
• What did the person learn between the first attempt and the hundredth?

Those questions tell us much more about creativity than the name of the tool.

I’m 83 years old. I’ve watched tools change throughout my lifetime. I’ve watched people resist new technology, eventually accept it, and then forget there was ever a controversy.

AI will undoubtedly change too.

But I don’t believe the arrival of a powerful new tool makes human creativity less important. I think it makes human creativity more important.

Because when almost anyone can generate something, the valuable part becomes knowing what is worth generating.

The machine can help water the seed. It can help it grow. It might even surprise you with where one of the branches goes.

But somebody still has to plant it.

And that part is still ours.

Let’s Have a Conversation:

Do you feel threatened by the AI expansion? What is AI to you – a tool or a human substitute? What are your fears in regards to AI? How do you make sure AI remains just a tool and doesn’t replace your creativity?