I need to start with a confession: this blog was never supposed to turn into whatever it is now. Truly. Honestly. I envisioned atmyboat.com as a fun, breezy chronicle of me learning to sail Our Time, my charmingly stubborn 1994 Monterey. I imagined writing about sunsets, minor upgrades, and the occasional “oops, I forgot to untie the stern line again” moment. You know — wholesome nautical chaos.
Instead, I somehow ended up building a marine AI operating system, inventing a methodology to control hallucinating language models, writing an economics paper, and creating a bow‑camera‑to‑AIS system that basically gives my boat superhero vision. I don’t know how to explain this except to say that I took one wrong turn on the internet and woke up in a completely different genre of life.
The moment everything went sideways happened on Lake Erie — because of course it did. There I was, drifting helplessly in the middle of the lake like a confused crouton in a giant bowl of soup. The culprit? An ancient cigarette lighter that wasn’t on a fuse. One tiny component, older than some of today’s influencers, decided to retire permanently. And with it went every electronic system on the boat except the Garmin and OpenPlotter, which I had hardwired to the house battery like a man who has watched too many episodes of Doomsday Preppers.
I had the paper manual, but it was about as helpful as a fortune cookie in a hurricane. So I turned to the OpenMarine forums, hoping for wisdom, guidance, or at least a sympathetic “been there.” Instead, the responses were… let’s call them “emotionally challenging.” So I pivoted to AI. And that’s when the universe said, “Oh, you wanted help? No. You’re getting a quest.”
I started with ChatGPT. Then Claude. Then Claude Code. And suddenly I wasn’t troubleshooting my boat anymore — I was debugging the hallucinations of large language models like some kind of digital therapist. I invented the AAO‑Methodology to keep the AI from wandering off into fantasy land. I wrote an economics paper about it. I submitted it to SSRN. I spent what can only be described as a “small but emotionally significant fortune” on API calls. At some point I realized I had accidentally become the world’s most reluctant AI researcher.
Meanwhile, my boat — the original star of this blog — was sitting there like, “Bro, I just needed a fuse.”
But no. I was already too far gone. I built signalk-forward-watch, a system that turns my bow camera into AIS markers so I can detect floating objects like I have radar but without the radar price tag. Then I added a weather radar that updates every 30 minutes. Then I built d3kOS, a full marine operating system with voice control, an AI first mate, and more features than a 1994 Monterey has any business hosting. I didn’t want to make another version of OpenPlotter, but apparently destiny had other plans.
Now I’m staring down the barrel of releasing version 1, updating the website, finishing the mobile companion app, integrating with ActiveCaptain, and preparing the actual physical boat for the season. I’m one project away from becoming a full‑time marine tech startup by accident. This was supposed to be a hobby. A fun little blog. A place to laugh about my misadventures, not a place to document the birth of a nautical AI ecosystem.
But here we are.
I promise — truly — that this blog will return to its original purpose: easy reading, lighthearted stories, and the ongoing saga of me vs. my boat’s electrical system. But first I need to finish releasing the software I never meant to build, update the website I never meant to expand, and prepare the boat I definitely meant to sail but somehow turned into a floating beta‑testing platform.
Stay tuned. The season is coming. And if history is any indication, something is absolutely going to go wrong in a way that is both educational and deeply, deeply funny. And I’ll write about it here — assuming I’m not drifting in Lake Erie again, clutching a paper manual like it’s a sacred text.
Leave a Reply