I've got a bunch of, let's call them, philosophical intuitions and positions that I got from reading this book. I've been meaning to do at least one post explicitly about Hierarchical Bayesian inference as a way to express what it is that keeps banging through my head.

First, though, I want to get Venture up and running to actually implement and train such a model, in progress, to make sure that my thoughts about the way it "should" work accord at all with how it actually does work.