From: dohduhdah@yahoo.com   
      
   Op 24/04/2024 om 23:08 schreef Stefan Ram:   
   > Richmond  wrote or quoted:   
   >> Some theorize that the purpose of consciousness is to allow us to do   
   >> psychology, i.e. to have some insight into what it is like to be someone   
   >> else.   
   >   
   >    A simple AI chatbot can often do a pretty good job at answering   
   >    psychological questions.   
   >   
   >    But the "insight" you mentioned is important in the following   
   >    way: We can experience pain. So we also see it as immoral to   
   >    inflict pain on others. However, we don't see it as immoral   
   >    to hit a wooden table, because we think it can't feel pain.   
   >   
   >    And that's consciousness - when we say it's immoral to hit   
   >    something, we're assuming it has consciousness and can suffer.   
   >   
   >    So, it's not the same as being able to process information.   
   >    An electronic circuit or a digital computer, they can both   
   >    process information just fine. But we don't think they're   
   >    actually feeling anything when they do it.   
      
   But the capacity to experience feelings is potentially not a necessary   
   condition to ascribe consciousness to an information processing system.   
      
   I think it's sufficient if we can argue that the system has some degree   
   of awareness of its situation, enabling it to navigate complex   
   situations successfully.   
      
   We can think of a minimal context where consciousness plays an essential   
   role, like learning or playing the game of go.   
   Some activities, like breathing, can be done without consciousness being   
   involved (like when you breathe while you're sleeping without dreaming).   
   Other activities, like playing go are unlikely to occur without   
   consciousness being involved, because it involves complex decisions   
   regarding the moves to choose to maximize the chances of winning the game.   
      
   An AI system might be conscious of what's at stake as it plays go in the   
   sense that it will attack your weak groups and defend its own weak   
   groups if you try to kill them. In fact, we can argue that these AI   
   systems have a more advanced conscious state than humans, because they   
   can easily beat the best human players (having learned the game from   
   scratch without any human knowledge as input besides the rules of the   
   game). So it stands to reason that they have an ability to discern   
   subtle patterns in the game that even elude the best human players,   
   enabling them to better anticipate how the game is likely to progress   
   depending on the moves they choose and possible ways their opponent   
   might respond.   
      
   The AI system doesn't have any emotions regarding the choices it makes   
   or the outcome of the game, but it does ascribe value to moves in the   
   sense that it preferentially plays moves that make it more likely to win   
   the game.   
      
   In the context of the game of go, consciousness is restricted to the   
   possibilities implicit in the rules of the game and other aspects are   
   irrelevant (like being conscious of the location where the game is being   
   played or being conscious of what you had for dinner a couple of days   
   ago) because they don't affect the outcome of the game under normal   
   circumstances.   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)