On Mar 24, 7:50 pm, jonesrob@emporia.edu wrote:   
   > On Mar 24, 2:39 am, "Randolph M. Jones"  wrote:   
   >   
   >   
   >   
   > > jonesrobemporia.edu wrote:   
   > > > Hofstadter has said (Godel, Escher, Bach... Basic Books,   
   > > > 1979) that "below every tangled hierarchy lies an   
   > > > inviolate level." But is this true? Over the past few years   
   > > > I have been developing my Asa (Autonomous Software   
   > > > Agent) AIs. At first Asa was a pair of neural networks   
   > > > (Trans. Kansas Acad. Sci., vol. 100, pg 85, 1997), later   
   > > > I modified it to employ case based reasoners (Trans.   
   > > > Kansas Acad. Sci., vol. 107, pg 32, 2004) and most   
   > > > recently a hierarchical structure (Trans. Kansas Acad.   
   > > > Sci., vol. 109, pg 159, 2006). The programming language   
   > > > was also changed along the way and I no longer run Asa   
   > > > on the original hardware.   
   >   
   > > > Now you may not know this but I am a flesh and blood   
   > > > living and breathing human being.  But just suppose that   
   > > > I too was an AI, running and modifing Asa. Couldn't two   
   > > > intelligent agents run and modify one another?  And   
   > > > modify whatever they wanted, backing up if and when   
   > > > their modifications caused a systems crash.   
   >   
   > > > Now perhaps the laws of physics could be held to be   
   > > > the "inviolate level" but I don't think that is what Hofstadter   
   > > > had in mind.  And I suppose we shouldn't let the agents'   
   > > > value system be too easily self-adjusted (Trans. Kansas   
   > > > Acad. Sci., vol. 107, pg 32, 2004) either.  But values   
   > > > can change if not too easily and too quickly.   
   >   
   > > You say that you and ASA could "run each other".  But the question is   
   > > "What is each of you running on?"  Somewhere down there, there has to be   
   > > a machine (mechanical, electrical, biological, molecular, whatever...)   
   > > on which things are "running".  That is the inviolate level.  For some   
   > > kinds of "programs", the laws of physics would be exactly what   
   > > Hofstadter means by an inviolate level (in my opinion)...but most   
   > > "programmable" machines would be somewhat higher (at least at a   
   > > molecular level, for example, like DNA transcription).   
   >   
   > I did change computers.  In principle at least I could have   
   > changed from an electrical PC to a mechanical Babbage machine.   
   > And from DNA one might change to RNA, etc.  I guess we must   
   > all use the same atoms to build up our machines? Or, I   
   > suppose, at least antimatter will work if not neutronium.   
   > And of course we are not at the bottom of physics yet   
   > either.  A Babbage machine follows Newtonian laws but a   
   > relativistic machine is possible.  With flight near the   
   > speed of light one can effectively speed up calculations.   
   > What might string theory bring?  Will our laws of physics   
   > always be changing too?  If so there goes your inviolate   
   > level. Will anything in science ever remain fixed?   
      
   I think you are misunderstanding what "the inviolate level" is.  For   
   example, an inviolate level would not change just because our   
   understanding changes of how the universe works, as you seem to be   
   implying here.   
      
   Any formal system must have an interpreter.  The formal system can   
   manipulate symbols and, at some levels, may even manipulate some of   
   its own interpretation rules.  But there is always some level of   
   interpreter machinery that the formal system has no influence over.   
   This is the "inviolate level".  For example, DNA can be transcribed to   
   produce proteins.  Proteins can copy and alter the DNA, or influence   
   the transcription process.  But neither DNA nor proteins can change   
   the physical laws directing how molecules attract to each other, fit   
   together, and interact with each other.  A computer program can   
   contain "self-modifying code", but it cannot change the hardwired   
   microprogramming from which all of the instructions are ultimately   
   implemented.  Or if you *do* give your computer the ability to change   
   the microprogramming, you are just changing where the inviolate level   
   is, because the program can't change the wiring structure or the   
   number of registers.  Or if you *do* turn your program into a robot   
   that can alter its own hardware, that's still subject to some   
   machinery (such as how electrons flow through the circuits) that it   
   cannot change.  When you do math, your manipulation of mathematical   
   symbols may change "how you think" in terms of how you think about   
   math, but it can't change the physics of the neuro-chemical processes   
   that implementing your thinking processes (this is independent of what   
   our current state of understanding is about how those processes   
   work).  This is what Hofstadter means about there always being an   
   inviolate level (in my opinion).   
      
   [ 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)