Note 01: R&D

I started using ChatGPT in earnest at the end of 2024. I had been excited about the technology for years, but it was watching my girlfriend Halle using it as a thought partner, beyond the simple searches and re-writes it was known for, that piqued my interest.

Last year, I sent 21,000 messages through the app—enough to be counted as a top 1% user—as I tested a range of ways to integrate it into my workflows in the studio and managing the building. As with any experiment, I wasted a good bit of time, but it became a helpful part of my daily life.

I wouldn’t have thought to write a blog post about it then, but things changed this past fall.

In September, I fell ill with mono and spent most of the rest of the year couch-bound. With nowhere to go, I started working on structured projects in ChatGPT, generative imagery in Midjourney, and “vibe-coding” with agentic tools like Codex.

These tools, and what they’re doing to my work, are the first time that my optimism about AI feels matched by its output. It feels grounding and magical.

So, I am starting this Research & Development publication alongside the Studio News to make brief notes about my experiments with AI.