Beyond breaking ciphers and building software, I've been making music for over 30 years. It's a creative outlet that balances the analytical side of my work—a different kind of pattern recognition.
The Beginning
I started making music in the early 90s with FastTracker—a tracker interface where you programmed music in a grid, typing note values and effect commands. It was basically coding with sound. Channels, patterns, samples, hex values. Pure numerical composition before anyone had a proper GUI.
That tracker mentality trained my brain for the same kind of thinking I use now. Patterns, sequences, precise timing, efficiency.
Multi-Instrumentalist (Self-Taught)
Guitar
Guitar came first. In school in Denmark, I was learning lead guitar parts from Finnish melodic death metal bands—Amorphis, specifically. "Drowned Maid," "Black Winter Day." Those mournful, beautiful melodic leads. I learned from tabs I found online in the early internet days.
Drums
Drums came next. Scandinavian metal was my classroom. I played covers of Dimmu Borgir, Old Man's Child—Norwegian black metal with blast beats, double bass, and tempo changes that would give normal drummers whiplash.
When "Mourning Palace" dropped in 1997, I was learning it on drums while it was brand new. The triplets on the double pedal when it speeds up—that was the killer part. I even attempted Nick Barker's insane drumming on "Kings of the Carnival Creation." That track is controlled chaos—carnival-from-hell energy at 250 BPM.
Joint pain eventually ended the drumming. Black metal is a young person's game. But the internal rhythm never left.
Piano
Piano I taught myself in a way that would make any music teacher cringe.
I couldn't read sheet music. Still can't. Traditional notation is an abstract encoding system—symbols on lines that require translation. My brain rejected it.
So I read MIDI files instead.
Load a MIDI file, watch the piano roll—literal visual representation of what keys to press and when. Same information, different visualization. One that actually made sense to my brain.
It's the same mental model as guitar tabs. Tabs show you exactly where to put your fingers. MIDI shows you exactly what keys to press. No translation layer. No abstraction. Direct mapping from visual pattern to physical action.
The Repertoire
Learned from Movies and Cartoons
My piano education came from an unconventional curriculum:
The Paradox
Here's the funny thing: I can play Hungarian Rhapsody, Moonlight Sonata, and Appassionata.
Ask me to play "Happy Birthday" and I'd struggle.
I never learned "piano." I learned those specific pieces. Pattern-matched them directly from MIDI into muscle memory. No theory foundation, no scales, no "basics"—just straight to the thing that was interesting.
Same pattern as everything else in my life. Skip the tutorial. Go straight to the final boss.
The Chorus Problem
For decades, I made music but never "finished" songs. My Dropbox archive has tracks dating back to 2002—hundreds of ideas, all incomplete.
I recently realized why: I only ever made the chorus.
The chorus IS the pattern. That's the hook, the core idea, the thing that grabbed me. Once I cracked that... the rest felt like filler. Padding. The boring parts.
- Verse? Setup for the chorus.
- Bridge? Transition to chorus.
- Outro? Wind down from chorus.
My brain would go: "I solved it. I found the thing. Why would I pad it out?"
It's the same pattern across everything:
- Ciphers: Solve the interesting part -> move on
- Piano pieces: Learn the sick part -> the rest fades
- Code: Build the core feature -> maintenance is boring
I'm a chorus person, not a full-song person.
The Solution: AI as Collaborator
This is why AI tools like Suno have been transformative for my music.
I have decades of choruses—core ideas, hooks, the "interesting parts." Suno can help build the verses, the arrangement, the full structure around them. It handles the parts I always abandoned.
Same with technical writing. I have decades of infrastructure knowledge—MongoDB, Docker Swarm, NGINX, systems that run billion-dollar companies. But writing it all out? That's the verse. That's the arrangement. That's the boring part AFTER you've already solved it.
Claude helps me get that knowledge out of my head and onto paper. The expertise is mine—10 years of 100% uptime, battle-tested production systems. AI just does the arrangement work so I'll actually ship it.
People ask "did AI write that?" They're missing the point entirely.
The knowledge is mine. The experience is mine. AI is just the tool that makes the chorus into a complete song.
Code by Day, Music by Night
I developed and maintain Kactuz Music—a music streaming platform built with my own open-source audio player library, wavesurf.
My friend Michael, known as Kactuz, is an EDC Las Vegas Discovery Project winner. What started as casual jam sessions evolved into a genuine creative partnership.
Building wavesurf was the same pattern as everything else: I had a problem (audio players fighting each other on a music site), I solved it, then I packaged the solution and shared it. Eat your own dogfood. Use what you build.
The music side and the code side aren't separate. They're the same brain doing the same thing—finding patterns, building systems, solving problems.
Just different outputs.
Production Timeline
30+ years of making music. Different tools, same creative drive.
Today's Output
On any given day, I might:
- Ship a viral technical guide (441K+ Reddit views this month)
- Publish an npm package
- Remaster a track from 2003
- Update multiple websites
- Produce new music
The LLI brain doesn't do one thing at a time. It runs multiple obsessions in parallel until each one is done.
Code by day. Music by night. Ciphers whenever the pattern won't let go.
"The same brain that made me a 'problem' turned out to be exactly what I needed."