🚀 go-pugleaf

On Fri, 26 Nov 2004 16:57:41 GMT, "EarlCox"
<earlcox@earlcoxreports.com> wrote:

>Well, if I look around at our world today, I would say that the human mind
>has accomplished a great many things, from building vast cities, developing
>a science that can explain the physical world in both its microscopic and
>macroscopic forms, to sending manned spacecraft to the moon and unmanned
>exploratory spacecraft to other planets throughout the solar system, to
>evolving a mathematics that can represent and work with highly non-linear,
>dynamic, multi-dimensional spaces, to writing great works of literature, to
>visualizing and manipulating single atoms, to creating a world wide
>communications network, to designing and mass producing amazing small and
>powerful computer-based machines, to ... well the list is impressive. yes,
>we still have our reptilian, xenophobic, tribal behaviors that may one day
>destroy the environment and plunge us into terrible wars. But we are not
>dumb, we are just, in a swarm intelligence way, constrained by our emotional
>and biological attachment to a world of "fight or flight" that has mostly
>disappeared.

   Not so long as the neighbors want to strong-arm you, steal
   your stuff, beat you senseless or do you in .... our main
   opponents in this world have always been EACH OTHER more
   than "nature". Keeps us sharp ... and, as a side benifit,
   has seeded most of those civilized accomplishments you
   brag about. 

>Computers are not only dumb they are just collections of circuits.

   So far. 

>Without
>software they are nothing.

   They COULD be made as modifiable firmware ... just
   as our brains are.

>It is not that our computers are dumb, it is our
>understanding of intelligence and self-evolving intelligent systems that is
>flawed.

   Missing. There's a BIG gap. We're OK on the macroscopic
   end of things ... psychology, sociology ... and we're OK
   on the microscopic end of things ... neurons, synapses,
   neuro-chemistry ... but we have been unable to ascertain
   how all the little bits and pieces go together to create
   the 'macro' result. Could take decades, maybe centuries,
   to understand ourselves with THAT degree of detail. 

>The software that will eventually replicate this process need not be
>deterministic or even programmed in the traditional sense. Today some
>advanced artificial life and cognitive scientists are beginning to realize
>that you cannot program a computer to be intelligent, you have find a way of
>evolving intelligence.

   Exactly. It's the only practical way - short of literally
   simulating someones entire central nervous system down to
   the molecular level ... which would probably run kinda
   slow. 

   In any event, EIs will surely have to self-evolve. We
   create the proper environment ... akin to a newborn
   brain ... provide an appropriate simulated environment
   and then let e-'nature' take its course. E-bugs, e-dogs,
   e-cats first ... then bigger stuff. The good thing is
   we could get hundreds of 'generations' in a day. The
   bad thing is that any emergent EI probably won't be
   much "like us" - a veritable 'alien' instead with a
   vastly different set of priorities, a different way
   of thinking, a different emotional makeup. 

   Can we all just get along ? I doubt it. Nothing in OUR
   history suggests such a thing ... we couldn't get along
   with the neandertals. Any tiny 'difference' can be a
   trigger for xenophobic violence.