How marine radar builds a picture out of nothing you can see, and the discipline needed to trust it in fog, at night, or in busy traffic.

How echo sounders actually work, and how the same instrument the course teaches for depth safety doubles as a fish-finder on the water.

What a nautical chart is actually encoding, and how to turn chart symbols into a position and a course you can trust.

A walk through a small cruising boat's core systems — hull, rig, engine, and the systems that keep them all talking to each other.

How LLM architecture evolved from RNNs to Transformers to Mixture of Experts, told as a chain of problems and fixes, with one example sentence carried through every stage.

Notes from completing the Day Skipper sailing course — working through chartwork, compass theory, tides, and collision regulations as a new blog category.

Why output-based hobbies — the kind you build with, not just consume — are the only ones that compound over time.

Why fake work isn't a personality flaw but a structural outcome of knowledge work being nearly impossible to measure directly.

A personal essay drawing a direct parallel between how LLMs are trained — pretraining, fine-tuning, RLHF, emergence — and the stages of a human life.

How the learning rate — one number in gradient descent — decides whether a neural network converges smoothly or blows up, explained through the one line of code that sets it.