Why I Started Making Music With AI: The Gap Between Imagination and Execution

I loved music long before I ever used AI to make it.

Growing up, I played piano and saxophone, dabbled with guitar, and later picked up harmonica and bongos. Music was never foreign to me. The problem was execution. I could hear ideas in my head that I could not reliably perform, arrange or produce at the level I imagined.

That gap between imagination and execution is the real reason I started making music with AI.

AI Did Not Give Me a Love of Music

AI did not create my interest in music, replace years of listening, or suddenly make every idea good. What it changed was access. For the first time, I had a practical way to turn rough lyrical ideas, moods, melodic instincts and production directions into something I could actually hear, compare and revise.

Before that, I often felt limited by my own technical ability. I could imagine more than I could play. That is frustrating when music is already part of how you think and process the world.

The Early Experiments Were Messy

Before my current workflow, I experimented with different AI music tools and generated a lot of material. Most of it was not release-ready. Some ideas were interesting; many were disposable. That period taught me something I still believe: generating audio is not the same thing as making a finished song.

The useful work happens when I decide what the song is for, write or reshape the lyrics, choose a direction, reject weak generations, keep strong moments, revise structure and document what I actually want to release.

Why Suno Became a Turning Point in 2024

In early 2024, Suno became a turning point for me because complete-song generation became usable enough that I could iterate on ideas quickly. I could finally move from “I hear something in my head” to “I have several versions I can judge.”

That did not remove the need for judgment. It increased it. More output meant more decisions: Which vocal feels right? Which arrangement supports the message? Which chorus is memorable? Which version should be abandoned?

That shift changed how I saw AI music. The value was not that a model could make a song for me. The value was that I could direct a larger creative process that previously felt out of reach.

I Wanted Music for Specific Moments

Another reason I kept going was simple: I wanted music tailored to situations that mattered to me.

Sometimes that meant spiritual reflection. Sometimes it meant a morning track, a relationship song, something playful, something reggae-driven, or an idea tied to a larger Jack Righteous story. Commercial music does not have to satisfy every personal use case. AI-assisted creation gave me a way to build around the exact mood, message or purpose I had in mind.

That freedom mattered more to me than trying to imitate what was already popular.

The First Revenue Changed My Perspective

I did not start because I expected AI music to become an easy income stream. When a few releases began generating modest revenue, the important part was not the amount. It was the evidence that something that began as an idea in my head could leave my computer, reach people I did not know and create measurable value.

I also learned quickly that distribution, rights, platform policies, promotion and audience-building are separate problems from song generation. A tool can help create audio. It does not automatically create a sustainable music business.

What I Believe Now

AI is most useful to me when it expands what I can test without outsourcing the decisions that make the work mine.

I still have to decide the purpose, the message, the standard, what gets revised, what gets rejected and what is worth releasing. That human layer became more important, not less, as the tools improved.

If you want to see the method that grew out of these early experiments, read Human + AI Music: The Jack Righteous Creation Workflow. It breaks down the current process from purpose and source material through generation, selection, editing, documentation and release.