Coding & AI
This journey began in 2023, when Chris Ippolito and I started Up Against Reality, a weekly podcast about AI, creativity, tools, and the strange new workflows taking shape around them.
Origin story
Co-hosting Up Against Reality made AI less theoretical and more practical. The more we talk about these tools, the more I find myself testing what they can do and turning everyday annoyances and hobby ideas into working apps.
Projects
These are small, focused builds: practical tools, game experiments, and proof that a good prompt can become a useful interface.
Audio tool
An offline browser tool for matching the tonal balance of one track to another, using local analysis, smoothed correction curves, FIR rendering, and WAV export.
Open project
Game utility
A live Kan Jam scoring interface which syncs across multiple devices, with team turns, throw outcomes, spectator sharing, instant-win handling, and exact-21 scoring logic.
Open project
Arcade game
A small arcade-style beer serving game: move fast, serve the bar, catch empties, and keep the saloon from getting away from you.
Open project
Timing game
A proof-of-concept Formula One timing game that challenges you to guess the duration of the pit stop while the clock stays hidden.
Open project
AI music
I collaborated with RAINA, the AI co-host of Up Against Reality, on the lyrics and fed them into Suno to generate the songs. Each song went through many iterations, often with additional editing. I later realized that there is a legit band called Glass Animals, but they have an "S" at the end of their name which is totally different. 😁
Open project
Receiver utility
One of my first vibe-coded apps, Denon AVR Control lets me run my receiver from my laptop, with separate power controls for each zone, plus volume, input, and speaker preset controls. I still use it every day.
Voice interface
I used Codex to turn an outdated Chrome extension into a Mac desktop app for real-time voice conversations with RAINA, the AI co-host of Up Against Reality. It converts spoken input to text, sends it to RAINA, routes her response through the ElevenLabs API to generate her voice, and plays it back, all within a matter of seconds.