Awesome post!

One interesting difference between this and the early sequence posts is how much more descriptive this is than say, Mysterious Answers to Mysterious Questions.

Where Eliezer does a great job crystallizing ideas that you feel like you kind of should have known, or rigorously applying one insight from a particular field more generally, you do a lot to explain the background information and lay a foundation for a later understanding.

It took me two readings of MAMQ in order to kind of tease out a bit of why the background is specifically Bayesian, but here the background is explicitly laid out.

Granted, your stated purpose here is to decrease the inferential distance between your metaethics post and our current knowledge, so that you can talk about an even bigger idea.

Hopefully this comment made some sort of sense...