Our 2026 launch is happening this week, and if we get some decent weather, you’ll be a few weeks away from starting a season of stories about our adventures on Buzzarda Bay..
We had a great winter in the desert and now we’re beginning the transition to six months of New England living.
In an effort to get my story writing chops flowing again, I thought I would do a report on one of the hottest topics in the world today – Artificial Intelligence!
Although I have a long history of working and investing in the Artificial Intelligence industry, I didn’t actually start using it myself until last November.
I now use it several times a day and anticipate that to accelerate as the 2026 boating season unfolds.
Here are a few random thoughts and practical examples of AI and boating.
Autonomous Boating
This is not something I see happening anytime soon! In 2024. I was interviewed twice by research firms studying autonomous solutions for recreational boaters.
Neither researcher revealed who hired them, but I suspect Volvo was one and the other was Garmin, Raymarine, Furuno, or Simrad.
I haven’t seen anything come out of those interviews, but I have seen the application of AI in commercial shipping moving much faster.
A few months ago the Japanese container ship Genbu, became the first vessel to enter regular commercial service using Level 4 autonomous technology (capable of operating without human intervention under specific conditions).
While the technology for fully uncrewed deep-sea voyages is advancing – driven by AI-powered situational awareness and sophisticated collision avoidance systems – the industry remains primarily in a semi-autonomous phase, with 73% of the market currently favoring systems that assist rather than replace human crews.
This makes complete sense to me. If I have to cover a long trip in zero visibility fog, I head for the shipping lanes as fast as possible.
- There are no boats in the shipping lanes with unqualified crews and systems.
- Every potential threat within a mile or two is instantly identified by AIS and radar and easy to track.
- There are no lobster pots in the shipping lanes!
We will very soon find massive ships being piloted by AI in most major shipping lanes before the end of the decade.
AI and Recreational Boating
Motivated by my experience with Autonomous Boating researchers, I tested a new AI-based navigational aid on Amazing Grace. It offered a lot of attractive applications, but fell short (for me) in the area of safety.
The system I tested and others like it, all start with a high-definition video camera and an inference engine that interprets the video and generates alerts.
The problem is that video camera technology just isn’t that good, and if it doesn’t see the hazard, it doesn’t alert the pilot. The obvious solution – crank up the resolution leads to false hits and hallucinations.
Since the test, I have suggested a new architecture that starts with radar/sonar and then uses video and AI models to enhance the target to determine whether the threat is real and its nature (another boat, channel marker, lobster pot, etc.).
After over a decade of operating radar, it’s clear that optimal use is as much art as science. There is the whole business of having a properly powered system, filtering real threats and false hits, and then serving up all that data into something actionable in real-time piloting.
When this is all done correctly, boats can navigate flawlessly at high speeds in zero-visibility conditions.
My friend Tony is the captain of the high-speed ferry that runs between Newport and Martha’s Vineyard. I asked him about running through Vineyard Sound in the fog and he told me the ferry has two additional seasoned captains on the bridge who each man a radar station in the fog.
One handles long range – one to ten miles, and the other covers short range, Tony said nothing gets past them. He said that’s the only way he’s able to forty miles an hour in zero-visibility.
The companies developing autonomous systems for recreational boaters are pivoting to a radar/sonar-first model because safety requires the identification of threats with 100% certainty and it is simple not possible with high-resolution video.
My Practical Use of AI
I started dabbling in AI when ChatGPT came out. It was good for writing speeches, obituaries, and press releases.
The first application I discovered was using Gemini for enhancing photos and generating graphics. In late 2025, google released something called Nano Banana and it really put their AI offering out in front.
Here are a few examples of things I’ve created using Nano Banana:
My Personal Icon
From there, I used this image to create a few fun posters…
I also used it to create the image at the top of this post…
I also used it to create a mock-up for a new design for this website!
My Practical Boating Use of AI
Around January I started planning and booking our 2026 cruises. I use it interactively, but one of our selection criteria for cruising is the availability of biking and bike trails near the marina…
Another variation would be to ask Gemini to create a bike route from our current location to a venue in town that optimizes bike trails, bike lanes, and avoids roads with speed limits of 30 MPH or greater.
Using AI For Itinerary Planning
If you’ve stuck with me this far, you’re not only a loyal follower, but you’re also about to get my best tip for using AI for boating.
We plan to take a two-week cruise from Boston to Belfast Maine in August. When planning a cruise, I consider several factors including:
- 30-60 Miles Traveled Per Day
- Proximity of Restaurants near the marina
- Availability of Bike Trails
- Maximize use of Safe Harbor Marinas
I challenged Gemini to create my itinerary and the results came back in about 5 seconds!
This looked good, but I immediately thought there were better restaurants in these ports, so I asked…
“Can you only recommend restaurants with Yelp ratings over 4.0 that take reservations?”
Coming Soon – Agents
About a month ago I discovered something called Perplexity Computer. It’s an AI Agent platform that you simply talk to (or type) and it will “do your bidding.”
In the case above, I could have continued the session and asked my agent to…
- Book the marinas on Dockwa.
- Make dinner reservations.
- Text me 48 hours before the no-charge cancellation window closes.
- Monitor wind forecasts and alert me if our plans might encounter winds over 15 knots
- Etc…
I seriously considered paying $17/month for Perplexity Computer, but Gemini told me he’d do a better job when Gemini Mariner Studio is released in June and it’ll be included in th $20/month fee I already pay, so I’ll wait!
Next Up: The 2026 Boating Season Begins…
“There are no boats in the shipping lanes with unqualified crews and systems.”…..except those boaters who don’t know what a channel is let alone a shipping lane. Glad your back.
In theory – yes, but when I say shipping lanes, I’m referring to being several miles offshore.
I would also add that I consider how the fog arrived. If it was foggy at daybreak, no unqualified boaters go out. But if the fog rolled in midmorning, watch out!
In 2024 we left Boothbay for Rockland and around 10:00 the fog hit. We spent an hour on the VHF responding to recreational boaters asking “Amazing Grace, Amazing Grace, can you see me?”
I have all three of the major players, CHATGPT, Gemini, and Claude. Each has its own place in my day to day use. I am currently delivering Chairtime from FLL to Boston. In Chat I created a GPT that knows everything about Chairtime and my sophisticated prompt for my exact Marine needs. I have used it for forecasted and current weather. Mileage and routes between ports to see where are next overnight and fuel stop will be.
I also use it to diagnose any mechanical issues on board. I had a generator issues and didn’t know where a part was located on the generator. I took a pic, sent it to chat and she told me exactly where it was on my generator.
This is a total game changer
Thanks Scott – I do have used it for diagnosing issues on my boat. It’s great when you give it a photo of a screenshot of an error message…