From: nospam@example.net   
      
   On Fri, 26 Apr 2024, Richmond wrote:   
      
   > D  writes:   
   >   
   >> They have no will of their own. Should they exhibit that, I think they   
   >> would quickly pass the turing test at least. Actually, in order to   
   >> pass the turing test, they probably would have to be "dialed down" in   
   >> order to not give perfect answers too quickly, then again, that is   
   >> also the charm of the turing test.   
   >   
   > The Turing test as far as I remember was intended to test if something   
   > could pass off as human. But I think that would be a useless A.I. as we   
   > already have more humans than we know what to do with.   
      
   I think it was meant as a test for artificial general intelligence that   
   would reach the human level. The idea being that if you cannot tell the   
   difference, then we can assume it is conscious and intelligent.   
      
   When it comes to having more humans than we know what to do with you are   
   missing one important thing.   
      
   Humans require salaries, food, they have limits on their time and the   
   tasks they are willing to do.   
      
   With AI:s a lot of those limits would go away, and in my opinion, would   
   lead to a golden age of post scarcity.   
      
   Humans would see what over time, by moving from a 2 day weekend, to a 3   
   day weekend, to a 4 day weekend. And at some time the whole concept of   
   work would shift to include human quality work only such as crafts,   
   artisanship, service, psychology, sports, competition, music, service etc.   
      
   My opinion is that mean while we need a lot of research into positive   
   psychology, to prepare people to find meaning outside of labour. We need   
   to develop techniques to help people find intrinsic meaning. If we don't   
   do that, we risk catastrophy when extrinsic meaning is taken away from   
   people.   
      
   Today we do have plenty of people with intrinsic meaning who do not need a   
   job to find meaning. We just need to understand how they do that, how to   
   encourage that, and perhaps to incorporate that in the raising of our   
   children.   
      
   > The question is, could an A.I. be conscious? I don't think a Large   
   > Language Model can be, but Google is threatening to release something   
   > which is capable of reasoning and planning. What do they mean by that?   
   > Could it plan something which has never been planned before?   
      
   Doubt it. I think it is mostly marketing. In terms of LLM:s I'm with you.   
   I don't think LLM:s will be "the" technology. Perhaps the break through   
   will come by combining different subsystems of which LLM:s will be one?   
   Imagine LLM, OCR, image generation, chess systems etc. as different   
   specialized "brain centers" and what's missing is will and an integrating   
   function.   
      
   > Can you tell if another human being is conscious? or do you just assume   
   > it.   
      
   Of course we all assume it. There is no way to conclusively prove the   
   subjective experience of another persons consciousness. We have to, by   
   nature, rely on proxies such as language, words, behaviours, brain scans   
   etc. So that is why I am a fan of the turing tests. If it passes, as far   
   as I can concerned, I will ascribe consciousness to it.   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)