occasional essays. mostly data-science-adjacent.
On typesafe ai's jev, a model that answers with probabilities instead of prose, and why most AI engineering tooling lives on the wrong side of the black box.
On AI tutors, the outsourcing trap, and a learning system whose one unbreakable rule is that the struggle stays mine.
Why ranked stopped feeling good: zero sum ladders, positive sum hobbies, and the slow migration between them.
A meditation on pattern recognition, and what physics might have to say about how to live.
Why fundamentals still matter in the age of agentic AI.