= 27d567bc2d48d9f40e9ccc20ea370b15)
ArisKatsaris comments on A sense of logic - Less Wrong
You are viewing a comment permalink. View the original post to see all comments and the full post content.
= 783df68a0f980790206b9ea87794c5b6)
Loading…= b715d02e85f7b3b4ae65ca3d3e450868)
Subscribe to RSS Feed
Powered by Reddit
= f037147d6e6c911a85753b9abdedda8d)
You are viewing a comment permalink. View the original post to see all comments and the full post content.
Comments (269)
It is. It's not natural to wonder if the first cause is a complex structured intelligent being, because such complicated and internally correlated structures demand simpler preceding causes of which to be the effects, for if we try to model the structure as uncaused we have unexplained internal correlations, which is a no-no in causal graphs.
If you then start making special pleading excuses about an intelligence that you predict using a complex structured internally correlated model but which you claim to have no structure so that you can pretend it's simple even though you can't exhibit any simple computer program that does the same thing, it's really unnatural - not just physically unnatural, but epistemically unnatural.
I'd like to taboo the word "natural" here. Do you guys mean 'good and reasonable'? Or do we mean 'typically occuring in human societies'? Or something else entirely?
My reduction-proposal:
A "natural" hypothesis is one with high probability. A "natural" question is a query regarding the cause(s) of a low-probability observation.
So, in this exchange, byrnema pointed to a particular low-probability observation (the abundance of causal structure in the world around us), and Eliezer responded by noting that the proposed explanation (a complex first cause) has low probability, even conditioning on the observation.
To put it in even simpler terms: Bayes's theorem says P(H|E) = P(E|H)P(H)/P(E); byrnema said: "P(E) is small!"; and Eliezer said "Oh yeah, well P(H) is tiny!"