home bbs files messages ]

Forums before death by AOL, social media and spammers... "We can't have nice things"

   comp.ai      Awaiting the gospel from Sarah Connor      1,954 messages   

[   << oldest   |   < older   |   list   |   newer >   |   newest >>   ]

   Message 1,742 of 1,954   
   Andrey Gavrilov to Ondra Zizka   
   Re: Is there any "general" AI theory cov   
   12 May 08 10:55:52   
   
   XPost: comp.ai.edu   
   From: andr_gavriloov@yahoo.com   
      
   "Ondra Zizka"  writes:   
   > "Andrey Gavrilov"  writes:   
   > | "Ondra Zizka"  writes:   
   > [[MOD: Snipped, see <4826887f$1@news.unimelb.edu.au>]]   
   >   
   > Not only the concept of society was an inspiration.   
   >   
   > The "general purpose" (groups of) neurons in the brain also must   
   > specialize for some task, and then they represent the old knowledge.   
   > Then there must be a mechanism which ensures that their specialization   
   > will not be lost, and more, that they will be found and used to solve new   
   > tasks.   
   > This is probably done by "random" trials of using them to solve it, and   
   > depending on the result of the action, their association to a group   
   > solving the new task is strengthened [or reduced] (e.g. through the   
   > emotions concept).   
   >   
   > The way how brain works may not be the only possible solution and we   
   > should not stick to it, but at least we know one working concept.   
   > This may not be the way how brain works, but sound reasonably.   
   >   
   > So, this business society model is an attempt to create a principle that   
   > could work on similar principles: allow specialization, reuse specialized   
   > units for new tasks, discard units that do not contribute to the solution   
   > of a task for other use. Maybe I should not have named it "business   
   > society".   
   >   
   > By the way, it seems that in special case it can form a neural network and   
   > backpropagation.   
   >   
   > I know that any concept I think out is likely to be at the level of 50's   
   > research; I am just a student after all :-) That why I asked for comments   
   > and references to the results of the research already done. Thanks for   
   > them.   
   >   
   > Ondra   
      
   I just tried to say in my comments that business or social models are not   
   good for usage in artificial itelligence because ones are more uncertain and   
   not estimateable than our mind.   
   Nevertheless may be you will achieve any interesting results.   
   Best wishes!   
      
   Andrey   
      
   [ comp.ai is moderated ... your article may take a while to appear. ]   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)   

[   << oldest   |   < older   |   list   |   newer >   |   newest >>   ]


(c) 1994,  bbs@darkrealms.ca