X-Message-Number: 14958 Date: Sun, 19 Nov 2000 10:42:25 -0800 From: Lee Corbin <> Subject: Re: nature of identity In Message #14934 John de Rivaz wrote >Suppose you telephone a friend for a chat just as he is >sitting in the sun doing nothing much wondering what to do next. > >Suppose you telephoned the same physical person just after >his hard disk had crashed, he has loads of work to do and >it is a fortnight since he backed it up. > >Is he going to react to your idle chat in exactly the same way? No, certainly not :-) at least not most people >If in each instant you had a very sophisticated computer >avatar program that analysed the resulting conversation and >tried to create a simulation of the person on the other end >of the telephone do you think it would produce the same result? Supposedly it would be possible to have two relatively small programs as you suggest: one would always respond just as if it was very relaxed and wasn't thinking of anything, and an entirely separate program would always respond as if a calamity had just occurred. But we can posit that a very large program (either feeding off of a very large database or having such information internal itself) would be capable of either behavior depending upon circumstances. Most programs that I know about are very sensitive to "state". That is, it may be in one state that is a function of the input that it has just received, while at the same time it also responds to timers that go off which cause it to also carry out entirely different functions. One program that I work with actually handles sixty incoming telephone calls at the same time, keeping track of which user has pressed which DTMF keys, and which user has just spoken what words (we have recently added speech recognition to our system, purchasing the really heavy-duty software from another company and just adding it on to ours). Of course, the "state" of any particular human being at any particular time is vastly more complicated than any state of any program today, but some of us think that eventually programs' states will be as complicated as people's. Lee Corbin Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=14958