I’ve been slightly dissatisfied with my take on epsitemology recently. Reading Pirsig I find I’ve no real model thats being updated here. It’s not really that I’m trying to dissprove Pirsig, just learn something.
For instance, I don’t get the ‘endless hypothesis’ part. Surely there is a finite number of explanations for all phenomena.
I was thinking last night about that word quality. It’s very hard to pin down. In a general way it has something to do with the ‘good’ which in turn has some purpose being achieved. What is good is an attribute that works towards some goal stated or otherwise.
Probabilistic algorithm’s?
Gym-
I listened to a podcast with Mike Isreaelti. The logically appraoch to gym work and generally acheiving goals is key. I mean its obvious from a decision theoretic standpoint too. I would say the number one reason I hold myself back is body composition issues. I feel that if I ‘bulk’ I’ll put way to much fat on my legs.
I think it would be good to go through my past notes and see what worked and didn’t work if there even is such patterns. Good exercise in using grep too.
I think its important to accept that health is important to me. That I want to be doing the right thing but thats a function of my goals and also my wellbeing now.
Decision Making
Acts and states should be causally independent.
Acts and events branch into the combinatorial stuff it seems.
Once you leave the realm of certain outcomes, each decision choice is a lottery on expected outcomes.
Utility is the value of the outcome from the decision makers point of view
Research
One of the themes I think I’d like to highlight in write up is the usefulness of uncertainty analysis. How does it improve the information that an LCA gives in a decision making process. Is there a way to quantify this?
If an entity is trying to make a decision and has some set of priors across all outcomes. In words, if the entity wants to know how much time and money (call it resources) to put into a particular method of biofuel to jet fuel creation (say, HEFA). The worse case scenario is they put all their resources available into it and that it turns out to be useless (in this sense, doesn’t reduce co2 gas emissions, is this utility?).
If this entities desired goal is to maximise income, then outcomes need to be represented in this income term. If this entities goal is to minimise something like co2 emissions, then outcomes should be represented in terms of CO2 emissions. It makes sense to choose the latter for the moment even if the former is the motivation for the latter.
Presumably, as long as this measure of consequences satisfies some axioms it can be considered a utility.