Feeble Minded Homo Sapiens
On Mon, 27 Dec 2004, Rich Cooper wrote:
> Party of Citizens wrote:
> > On Mon, 27 Dec 2004, Rich Cooper wrote:
> >
> >> There is an article in Technology Review's web site
> >> on Mitsuo Kawato's project to build a robot as a way
> >> of learning how the brain functions.
> >
> > Why bother with simulating the feeble brain of homo sapiens? Why not build
> > a better brain for robo sapiens? If USM doesn't, its enemies will.
>
> Feeble? Erratic as the human mind is, its light years ahead of
> anything we can do now, with any technology we have, that's
> related to the pattern recognition capabilities, analogous reasoning,
> induction and abduction capabilities that even three year olds can
> do while standing on their heads.
>
> Just simulating the subjective capabilties of one robot could
> help us understand one more increment of how that "feeble"
> human mind works. Then we can try for the next increment.
> Its going to take a while.
Take the tests which typically go into a profile of mental abilities.
Compare the memory of RS with HS. Compare the logic of RS with HS, eg in
chess playing. Remember when RS identified the few faces with police
records out of tens of thousands at the Super-Bowl? What HS could do that
so fast? What HS can rotate a complex three dimensional object with such
accuracy and so fast?
Now what weights are you going to assign to the tests of the profile? Give
the weights which favour RS and you already have a machine with SHAI.
POC