This brings to mind the dimensionality of the space of possible actions.  I 
don't play the Trolley Game Nick wants to play because the space is absurdly, 
artificially small ... presented by sophists who care little about the real 
world.  The real world (usually) presents a very high dimensional space of 
possible actions.  And those actions are (usually) not atomic, but composite.  
So, when presented with a situation where computing the likely first-order 
consequences is too expensive, rather than wait it out, one can chop up the 
potential (composite) actions into their first-executed action-lets, then sense 
and go again.  Or, perhaps if you've done the pre-processing, you can take the 
action-lets that have been identified as "critical path".  If some action will 
be required *anyway*, common to a wide array of potential actions and 
consequences, then do that.

This sort of agility is more common than pure rationalists/analyists/idealists 
would have us believe.

On 07/11/2018 08:39 AM, Marcus Daniels wrote:
> There’s another type of paralysis that does not involve doubt but computation 
> cost in resolving or characterizing doubt:
> A soon as it takes more time to compute the likely first-order consequences 
> of various potential actions (even using coarse-graining or surrogate models 
> or whatever simplification) than just waiting for time to pass, action is 
> arbitrary.   One could have precise information about all the aspects of a 
> decision but putting it together into a recommendation for action could be 
> too hard.  One could probably even prove it is impossible in some cases.

☣ uǝlƃ

FRIAM Applied Complexity Group listserv
Meets Fridays 9a-11:30 at cafe at St. John's College
to unsubscribe
FRIAM-COMIC by Dr. Strangelove

Reply via email to