Page 81 - An Evening with Maxwell's Daemons
P. 81

The Machine in the Ghost

        they’ll invent a new mousetrap we hadn’t imagined possible in our
        wildest dreams.”
          “Now, really, Hydrargyrum,” said Fred Feghootsky, “you seem
        to be getting rather emotionally involved here. Granted, the battle
        for supremacy is a grand and relevant theme; and you certainly are
        proof that it touches a raw nerve in a society already bemused and
        bewildered by rapid technological change. But I would take to heart
        your comment about the well-worn plot paths involving ghosts in
        machines versus machines in ghosts: if we organisms are nothing
        more than meat, then, yes, the metal monsters could overcome us.
        If, on the other hand, some quality—call it emergent, or gestalt, or
        even  the  whole  nine  yards  of  dualism  and  independent  souls  or
        minds—is ours alone, then we would always have a possible means
        of defeating the cold, unblinking eye. Yes, I know: that means has
        already been invoked in the simple idea of our supposed ability to
        act irrationality when the machine cannot. But that is like a chess
        gambit the opponent cannot either anticipate or accommodate, and
        everyone knows there will never be another Grand Master capable
        of beating the computer.”
          “Well, what is your point?” asked Izzy. “An appeal to mysticism?
        That’s  functionally  the  same  thing  as  acting  wild  and  crazy  to
        confuse the rational mind. You are limiting AI’s knowledge of its
        creator. That presumes stasis, not a dynamic intelligence.”
          “Aha!” It was Leith Mauker. “You slid over something crucial.
        Evolution may favor the superior hominin brain, but even identical
        septuplets will have different thoughts. Humans develop over time,
        like the crystals Izzy mentioned, subject to all sorts of contingent
        factors. One of those influences might well be unconscious: I think
        it either is known or will be discovered that our selfhood depends
        on various feedback loops involving parts of the brain we cannot
        easily access. In our case, it is both protective, preventing overload
        in  infancy  and  childhood,  and  destructive  in  terms  of  distressing
        experiences  leaving  their  mark  on  consciousness—that  is,  self-
        awareness:  yet  another  reflexive  process  partially  submerged  and
        dimly  perceived.  Could  an  artificial  intelligence  have  an
        unconscious? Yes, in terms of unbreakable rules, the so-called ‘laws

                                       80
   76   77   78   79   80   81   82   83   84   85   86