The producer discusses the chaotic, single-take philosophy behind his new EP and why generative AI can’t replicate the human spark of a dorm-room discovery.
Flying Lotus makes music that feels like a conversation between machines, but its origin story is entirely human. He traces a pivotal moment back to a college dorm. “There was this guy who lived in the dorm, he was a great friend and an awesome artist and he actually made cool music with the computer, and it was like… ‘wait, you can do that?’” That revelation, the simple shock that creation was possible with the tools at hand, still informs his process. It’s a spirit of immediate, personal discovery that he finds fundamentally at odds with the current wave of generative AI.
His new EP, BIG MAMA, acts as a manifesto for that human spontaneity. Described as a chaotic speedrun, it’s a dense, five-minute blast through chiptune, breakcore, and jazz fusion. Crucially, it was built without looping a single bar. Each moment is a one-off, a decision made in real time. The result is a composition that feels alive and unstable, mirroring the flow of an improvised jam rather than a grid-locked construction.
This approach is his answer to the homogenizing threat he perceives in AI music tools like Suno. For FlyLo, the value isn’t in the perfectly generated loop, but in the flawed, idiosyncratic choice. The magic is in the “you can do that?” moment, a personal epiphany that software cannot simulate. His work has always existed in the space between programmed precision and organic looseness, and BIG MAMA pushes that tension to its logical extreme.
The EP is less a collection of songs and more a captured event. It’s the sound of an artist refusing to let the computer have the final say, instead using it to document a fleeting, complex mood. In an era leaning towards automated composition, Flying Lotus is championing the irreplicable spark of a single, unrepeatable take.
Join the Club
Like this story? You’ll love our monthly newsletter.
Thank you for subscribing to the newsletter.
Oops. Something went wrong. Please try again later.






