Worked on the first version of my self-experimentation protocols and tools. Basically, I’m going all professional, and I’m developing precise (as in, numerical answer and specific algorithm, not “that feels about right”) methods to do self-experiments. Mostly because I’m extremely ashamed and disgusted that I don’t have those already1.
I’m writing all of this up, of course, and I’ll publish drafts as soon as possible. This is literally top priority for me. The only reason I don’t have a draft out already is because I want to do one complete run so I see that I’m not missing anything, to demonstrate to myself that I’m not doing any handwaving. Then I’ll iterate to victory.
The idea is to do meta-empiricism: test empiricism, see if it works. Once I have some meta-data, some actual “look what awesome results that brought me”, beyond “that’s like that thing Eliezer was talking about”, I’ll actually advocate people do that, and shame2 those who don’t. If it doesn’t, I’ll drop the “rationalist” bit from my identity and become a bitter troll.
(Also, the Librarian is a true rationalist. Learn from him, and beware the alien, the mutant, the heretic.)