""" In a world that has no regularities at all, there is no benefit in trying to find system-level mappings between action and reaction because will just be different every time. Our friend Will is tasked with navigating this impossible space, but it is impossible as defined? If there are some regularities, conditional probabilities that can be tabulated, then Will can start to play the odds by learning the distributions that are observed together with different trajectories that may become evident as it steps into to the game. Maybe there are hidden variables that explain the apparently random generators? For example, ought there not be some prior state that can explain why the Will stepped into this game in the first place? Or do we assert, as the Free Will contingent do, that Will is above the fray? """
I read you as outlining why Will *ought* to get to work feeling for regularities (even if these regularities are simply local). But she *ought* to (and in theory *will*?) only in a world where she is *free* to do so. Regularity is ambiguous here: 1. local regularity: When I see a 'he' it is often preceded by a 't', and so Bayes gets to work. 2. global regularity: It will be the case that once all novelty has been generated, I will compressible structure. I agree that for agents with a choice (say), option 1 is an exploitable strategy even if we ultimately do not get to rely on option 2. Ah, opacity between worlds, was that David Lewis? I am not sure it is the kind of thing that is solved by a big enough parallel computer, especially if we mean a non-theoretical computer. Anyway, thanks for humoring me. -- Sent from: http://friam.471366.n2.nabble.com/ - .... . -..-. . -. -.. -..-. .. ... -..-. .... . .-. . FRIAM Applied Complexity Group listserv Zoom Fridays 9:30a-12p Mtn GMT-6 bit.ly/virtualfriam un/subscribe http://redfish.com/mailman/listinfo/friam_redfish.com FRIAM-COMIC http://friam-comic.blogspot.com/ archives: http://friam.471366.n2.nabble.com/
