Hands-on experiments exploring how AI, data, and autonomy translate into real product capability. These projects are exploratory by design—used to test assumptions, understand limits, and sharpen product judgment.
Data collection, training pipelines, and control-system experimentation for an autonomous RC car.
October 2025: “Interactive Imitation Learning” Previously, I had built Carbot to collect data while in ROV mode — capturing images, then processing them through a CNN in an attempt at behavioral cloning. It didn’t work well at all. Most of...
Read more →Carbot continues to grow. Now I have a full ROV mode available. It’s rather amusing because it really is an Android phone on top of a remote control car. The phone talks to an Aduino via BLE to control the...
Read more →
For the last several months I have started down a different path with Carbot. I’m working out what it will take to make it into an Remotely Operated Vehicle (ROV). The reason for this is pretty straight forward, In my...
Read more →Small analyses and modeling experiments focused on data quality, inference, and practical decision support.
Eventually Complete Digital Twin is a term I’ve been using with colleagues, and I realized I never fully explained what I meant. So I wanted to take a minute to break it down and explore the idea more clearly. This...
Read more →
I bought an Epic Pass like so many other people in Seattle. I have been renewing it for the last 3 years. Here in Washington, the only place you can use it is Steven’s Pass which is a fine resort...
Read more →Exploratory Data Analysis (EDA) is a crucial step in understanding the data you are working with before applying any models or drawing conclusions. It involves summarizing the main characteristics of a dataset, visualizing patterns, relationships, anomalies, etc. Below are the...
Read more →Rapid product experiments using AI-assisted development to test hypotheses and explore technical feasibility.
Lessons From 10 of AI-Driven Idea Spikes In late 2025, I ran a personal experiment as part of my 10-week goals: build a different idea every week. I started calling these Idea Spikes—a fixed amount of time allocated to create...
Read more →For this last and final spike of 2026, things got a little strange. I didn’t make a software product. I made a song. Even that feels generous—I didn’t compose it so much as prompt it into existence. I went to...
Read more →Quirks was a game that some friends and I dreamed up over a decade ago. It started with a simple observation: everyone has those weird little behaviors—those personality “tics” that make us who we are. Some people take endless notes....
Read more →