DrumZombieArtist Statement
How the music is made

A straight answer about my production process, what "AI slop" actually means, and why this isn't it.

I've been playing drums and keys for a long time. Barn lofts, farmhouse weddings, Virginia wineries, stages that rewarded feel over polish. That experience doesn't disappear when I open a laptop. It's the reason everything I make sounds like it was written by a human, because it was.

What's changed is the size of my studio. I used to need a band, a recording space, a budget, and a schedule that matched everyone else's. Now I need a blank page, my instruments, a DAW, and the right tools. One of those tools is AI. That doesn't make me an AI artist any more than using Pro Tools makes me a software engineer.

There's a lot of noise right now about "AI slop," and most of it is justified. There are people flooding streaming platforms with thousands of auto-generated tracks designed to siphon fractions of a cent from algorithms without a single human creative decision behind them. That's a real problem. It's also got nothing to do with how I work.

Here's the actual process.

The Pipeline

01
Step one
Write it. All of it.

Every lyric, every melody, every structural idea comes from me: blank page, no prompts, no generation. The concept, the story, the hook. This is where the song lives or dies before any technology touches it.

Human only
02
Step two
Track the foundation live.

I record myself playing the core elements: guitar, drums, and keys. That locks in the rhythm, feel, and tempo of the song. This is the skeleton. Everything built after this reports to it.

Human only
03
Step three
Bring in the AI session band.

Using Suno (paid Pro subscription, full commercial license), I generate synthetic vocal layers, harmonies, and additional instrumentation around my existing foundation. I direct it the same way a producer directs a session musician, with a specific result in mind. The AI doesn't write the song. It fills the wall of sound I've already designed.

AI as instrument
04
Step four
Rebuild it in Cubase.

The generated audio comes into my DAW, where I slice, rearrange, and strip out what doesn't serve the track. I pick up my instruments again: live piano, keys, percussion, and backing vocals. Then I layer them over the synthetic material to add organic texture, human imperfection, and analog warmth. Nothing leaves raw.

Human + AI composite
05
Step five
Mix, master, release.

Final mix and master to bring the live and synthetic elements into a cohesive, radio-ready composite. Distributed through Ditto under the DrumZombie artist name.

Human-directed finish

"If you strip the AI out of my tracks, you're still left with a hand-written, live-recorded song. The AI fills the wall of sound. It doesn't build the house."

What AI Slop Actually Is

The term gets thrown at anything that involves AI in music production, which is lazy criticism. The actual definition is narrower and more specific, and understanding it makes the distinction clear.

AI slop: extractive
GoalFlood platforms, siphon royalties from streaming algorithms
OriginText prompt, no human musical input
ProcessGenerate to upload. That's the entire workflow.
IPImitating existing artists' sounds without licensing
Human roleNone. Automation is the point.
ScaleThousands of tracks, zero curation
This workflow: creative
GoalWrite and release original music worth listening to
OriginHand-written lyrics, live performance foundation
ProcessWrite to track to direct to produce to master
IPOriginal compositions, paid commercial license
Human roleDirector, lyricist, musician, producer, editor
ScaleOne song at a time, heavy curation at every stage

The difference isn't which tool was used. It's whether a human being made creative decisions at every step, or outsourced all of them to automation.

The Honest Answer on Training Data

There are two separate criticisms that get blurred together constantly. Let me separate them, because they're not the same argument.

Criticism one: AI tools "sample" existing recordings. This is factually incorrect. Suno doesn't store or recombine audio clips from real songs. It works like a predictive synthesizer. It has learned the mathematical relationships between sounds and generates entirely new audio from scratch. No artist's recording is hidden inside the output. The analog is a synthesizer, not a sampler.

Criticism two: AI models were trained on existing recordings without artist consent. This one is real. Early AI music models trained on large public datasets, and many artists never agreed to be part of that process. That's a legitimate debate worth having honestly.

My position on it

Think of a musician who spent twenty years absorbing recordings by Waylon Jennings, Merle Haggard, and Chet Atkins. That musician has been "trained" on their work without compensating them. We don't call that musician a thief. We call that influence. The question is where the line falls when a machine does the learning instead of a human. That line is being actively redrawn by the industry right now.

2024
Major labels including Sony, UMG, and Warner filed suit against Suno over training data practices. The criticism had merit.
2024-25
Warner Music Group reached a commercial licensing agreement with Suno, moving from litigation to structured compensation. The same arc that turned Napster into Spotify.
2026
Suno is transitioning toward models trained on licensed, opt-in datasets. The industry is formalizing what was previously unstructured.

I use a paid Pro subscription under Suno's commercial terms, which grants me full commercial ownership of everything I produce. I'm on the side of this that's moving toward accountability, not away from it.

About the Range

DrumZombie is country-adjacent as a home base. I genuinely love the genre. The storytelling tradition, the directness, the pedal steel guitar. There's an emotional vocabulary in country music that cuts straight through, and that's where a lot of my songwriting lives naturally.

But it's not the ceiling. I'll put anything on DrumZombie that I find interesting and worth releasing. The throughline isn't genre. It's that I made it, it went through the full pipeline, and it's worth your time. If you follow this project, you're following a musician with range, not a genre jukebox.

Back to DrumZombie