Making Jazz Fusion with ChatGPT and Suno: My Experiment
By AudioPhil (Phil Reichert) · Published · Updated
ChatGPT supplied the text; Suno generated the audio; I chose the direction and the take to keep. This was a manually guided experiment, with separate tools and listening decisions between stages.
1. The inspiration: from drum breaks to style combinations
In Beat Magazine (October 2025, p. 13), I encountered a discussion of beat juggling that sent me down a musical rabbit hole. The appealing idea was that familiar recorded material could become something new through selection, timing and manipulation.
There is a useful historical distinction here. DJ Kool Herc is associated with extending drum breaks, an important foundation of hip hop. Beat juggling developed into a more elaborate turntable practice: manipulating recorded beats and their components to construct new rhythmic patterns. NAMM’s oral history of DJ Steve Dee credits him with developing the technique he called “The Funk”, subsequently known as beat juggling.
My experiment borrowed that spirit of recombination. It did not reproduce the turntable technique. I was combining descriptions of musical styles, rather than manually rearranging recorded beats. Could jazz fusion and an unexpected partner give a music generator a useful creative direction?
2. A staged prompt, rather than an autonomous DJ
My earlier article, Five Fresh AI Trends for Productivity, encouraged me to think about sequences of tasks rather than isolated requests. Here I used one compound prompt to ask for four things: explore combinations, select one, describe its sound and suggest a title.
This is best described as a staged prompting workflow. ChatGPT did not operate Suno, manipulate turntables or listen to the finished recording on my behalf. I transferred the text into Suno and assessed the music myself.
The original ChatGPT prompt
The wording records how I approached the experiment at the time. I would now replace “beat juggling” with “musical style combinations”, and ask explicitly for a reason behind the selection. “Most innovative” is a subjective preference, rather than a measurable result.
The working response
The following response supplied the candidate combinations and the selected direction. Its descriptions are creative proposals, not verified accounts of how the eventual audio was produced.
🎚️ 10 Beat Juggle Style Combinations
- Boom-bap × Lo-fi swing
- Gritty vinyl kicks over shuffled hi-hats and tape-worn snare ghosts.
- Trap × Latin percussion
- 808 subs blending with congas and timbales for polyrhythmic drive.
- Breakbeat × Afrobeat
- Chopped UK breaks sliced with Tony Allen–style syncopation.
- Jazz swing × Jungle
- Upright bass groove cut with high-tempo amen breaks.
- Funk × House shuffle
- Live funk drum loops re-quantised to 4/4 with syncopated claps.
- Soul break × Synthwave
- Warm Rhodes layers and 80s gated snares.
- Drum-and-bass × Bossa nova
- Brushed snares and double-time drums in silky counterpoint.
- Neo-soul × Glitch hop
- MPC-style beat cuts fused with bit-crushed micro-samples.
- Cool jazz × Boom trap
- Brushed kits and 808 sub rolls in smoky dialogue.
- Fusion jazz × Industrial
- Live cymbals and ride grooves over metallic percussion and modulated filters.
🌀 Most Innovative Style — Jazz Fusion × Industrial
This pairing captures both artistry and audacity: the improvisational phrasing of 1970s fusion re-interpreted through the mechanical grit of modern industrial rhythm. Think Herbie Hancock meets Nine Inch Nails at a late-night warehouse jam.
The drums are live-recorded but sliced into asymmetrical loops with syncopated ghost notes that constantly shift the pocket. Bass lines pulse through modular synth oscillations, trading bars with electric piano licks processed through warm analog saturation. Horns stab in call-and-response with distorted guitar textures, while atmospheric pads provide the illusion of smoke and stage lights. The beat juggling blends acoustic swing with robotic precision — organic phrasing dancing against quantised steel.
🎵 Three-Word Song Title
Chrome Blue Groove
The chosen combination was jazz fusion × industrial, with the title Chrome Blue Groove. Electric piano, syncopated drums and metallic textures gave me a more useful starting point than a request for something merely “jazzy”. The title also supplied a mood I could judge the music against.
3. Taking the description into Suno
I used Suno v5 with a Pro subscription and chose an instrumental. Removing lyrics helped me concentrate on the relationship between the style description, the arrangement and the groove.
These are the settings recorded for the experiment, rather than a claim that they are optimal or a description of the service’s current interface.
- Model:
- Suno v5 (Pro)
- Prompt Style:
- Custom
- Lyrics Prompt:
- Instrumental
- Style Prompt:
- Jazz Fusion × Industrial — This pairing captures both artistry and audacity: the improvisational phrasing of 1970s fusion reinterpreted through the mechanical grit of modern industrial rhythm. The drums are live-recorded but sliced into asymmetrical loops with syncopated ghost notes that constantly shift the pocket. Bass lines pulse through modular synth oscillations, trading bars with electric piano licks processed through warm analog saturation. Horns stab in call-and-response with distorted guitar textures, while atmospheric pads provide the illusion of smoke and stage lights. The beat juggling blends acoustic swing with robotic precision — organic phrasing dancing against quantised steel.
- Advanced Options:
- Weirdness 50 % | Style Influence 50 %
- Song Title:
- Chrome Blue Groove
How to read the style prompt: phrases such as “the drums are live-recorded” describe the requested aesthetic. I did not record a live drummer for this track, and the prompt does not establish the provenance of Suno’s generated sounds. Likewise, a request for sliced loops or modular synthesis does not prove that those production methods were used.
4. Listening and choosing a take
I exported two versions and listened to both. I preferred the take that seemed to give the drums more breathing room between phrases. The laid-back kit, lightly swung ride and bass movement were the features that held my attention.
The useful result was a recording I wanted to keep listening to. That is a personal listening judgement, rather than proof that the model followed every detail of the text. The style paragraph set a direction; it was not a score or a guaranteed production specification.
I published the recording through the Internet Archive so readers can compare the musical result with the instructions above. The embedded player retains the original media filenames, Chrome-Blue-Steel.mp3 and Chrome-Blue-Steel.flac; the track is presented here under the article’s title, Chrome Blue Groove.
5. What I would change in the next experiment
I asked ChatGPT to critique the original prompt. Its most useful observation was that the sequence was clear but the selection criteria were vague. Next time I would define what I wanted from the combination before asking the model to choose it.
- Define the musical contrast. Ask for a recognisable jazz-fusion foundation and one contrasting rhythmic or textural feature.
- Make the choice reviewable. Require a brief explanation of why the pairing is musically coherent, rather than simply calling it innovative.
- Keep the final style description concise. Prioritise instruments, feel and arrangement over decorative imagery.
- Record the listening decisions. Note which features appeared, which were absent and why one take was preferable.
The earlier critique also suggested exploring a broader range of rhythmic traditions. That is a useful creative invitation, but the model’s comments were not a systematic assessment of musical or cultural bias.
Try a revised prompt for a similar experiment
This revised prompt is a proposal for a future trial; it was not the prompt used to generate the recording above. Precise targets such as swing percentages, bar counts or mastering loudness would need checking in the audio. Putting them in a text prompt would not establish that the output met them.
Conclusion: the listening decision matters
A magazine reference led to a style combination, a style combination led to a prompt, and the prompt led to two recordings to compare. The workflow made it easier to move from curiosity to something audible.
What I would keep is the sequence: explore, choose, describe, generate and listen. What I would improve is the record of that final judgement. The prompts explain what I asked for; the listening explains what was worth keeping.
Sources and experiment notes
- Beat Magazine, October 2025, p. 13: the magazine reference that prompted the experiment.
- NAMM: DJ Steve Dee oral history, for the historical distinction concerning beat juggling.
- PBS: The Art of Turntablism, for background on Herc’s break-beat technique and subsequent turntable practices.
- Internet Archive recording collection: external hosting for the embedded audio.
Disclosure: ChatGPT contributed the musical brainstorming, style descriptions and prompt critique shown here. Suno generated the audio. The cover art is AI-generated. This article documents my experiment and listening choices; generated descriptions should not be treated as verified production facts.
