XPost: comp.ai.alife, comp.ai.games   
   From: Jochen.Fromm@t-online.de   
      
   The exact quote of Herbert A. Simon is   
   "A man, viewed as a behaving system, is quite simple.   
   The apparent complexity of his behavior over time is   
   largely a reflection of the complexity of the environment   
   in which he finds himself"   
   (from "The Sciences of the Artificial")   
      
   This is true. Simon was so fascinated by his   
   fundamental insight that he repeats it several   
   times in his book.   
      
   So without complex environment,   
   there will be no complex agents. But the   
   successors of the current Multiplayer games   
   will already have a very complex environment,   
   because the complexity of the environment arises   
   largely from the agents who live in it and their mutual   
   interactions. These games will come very close to   
   reality as we know it.   
      
   Complex artificial worlds have in fact a lot of advantages   
   for training up of an AI, compared to the real world.   
   You can do whatever you want, for example duplicate   
   and split your agent a hundred times, let it learn a hundred   
   things in hundred different contexts and places, and merge   
   the resulting hundred agents together in one agent again.   
   With the right algorithms, you can speed up learning   
   enormously.   
      
   And there are other advantages. You can see the world   
   with the eyes of the agent or from every other viewpoint   
   at every time. This possibility can be used for different   
   mechanisms and things, for example to improve learning   
   methods or visualizing the state of the agent.   
      
   [ comp.ai is moderated.  To submit, just post and be patient, or if ]   
   [ that fails mail your article to , and ]   
   [ ask your news administrator to fix the problems with your system. ]   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)