From: Eliezer S. Yudkowsky (email@example.com)
Date: Mon Mar 14 2005 - 23:37:52 MST
Psy Kosh wrote:
> Since advanced Bayesianisim/metalogic/etc is being thrown around
> lately, figured I might as well ask a possibly stupid question that's
> been on my mind:
> How does one train themselves to this? What I mean is yes, I am
> familiar with Bayes theorem, yes, I understand the idea, and if you
> have be a list of probabilities, I could juggle them around just fine,
> and would also know to do that instead of making some of the more
> obvious mistakes...
> But, other than those cases that one is explicitly handed the
> probabilities on a computronium platter, how does one actually train
> oneself to parse everyday life/thinking in terms of it?
> ie, lots and lots of incoming info, not obvious to me how to
> intuitively parse it into a form suitable for this type of reasoning.
> So... to those who are learned in the Craft, I ask: how does one train
> oneself to gain the appropriate intuitions? Does one practice
> meditating on venn diagram "mandalas"? (I'm only half-joking here, I
> really am stumped as to how to actually organize my mind for this)
> I appologise if this isn't really an SL4worthy, but since the subject
> seemed to be discussed lately anyways, and simply since I'd have no
> clue where else to ask this... :)
The first question to be asked is:
Have you read _Judgment Under Uncertainty_, _Heuristics and Biases_,
_Choices, Values, and Frames_, and some volume of social psychology
(say, the Blackwell Reader in social psychology)? And have you browsed
some Jaynes? Robyn Dawes's _Rational Choice in an Uncertain World_
makes for a good intro if you're not ready to plunge straight into
_Judgment Under Uncertainty_.
Unfortunately I have not yet gotten around to founding the Baisutsukai
Order of Solemn Rationalists, starting a Bayesian monastery, or even
founding a Bayescraft dojo - I have these other minor things on my mind
- so I think that's pretty much as far as it currently goes for aspiring
rationalists. Just do plenty of reading. The above will get you
started and tell you where to poke around for more.
-- Eliezer S. Yudkowsky http://intelligence.org/ Research Fellow, Singularity Institute for Artificial Intelligence
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:50 MDT