[ExI] The digital nature of brains (was: digital simulations)
steinberg.will at gmail.com
Fri Feb 5 21:15:32 UTC 2010
Neurons also encode information based on the relative strengths of
connections with adjacent neural structures, as well as in their
propensities towards different underlying currents. It seems like a lot of
you, perhaps due to the polarizing atmosphere of the land of GetSwobia, have
too easily discounted legitimate logic and mathematics.
Now--the Chinese Room is an absolutely ridiculous analogy and Searle should
be demeaned for it, as should anyone who takes it too seriously, because
semantics ARE NOT STATIC. I do not understand how anyone can begin to make
any arguments centering around this sort of ill-thought out, sophomoric
idea. It's as if Searle simply took the first thought experiment he could
think of, immediately deciding, without considering reality, that the rules
of understanding could be "booked"--simply an untruth. Really, the man
should have a second set of books which give him rules for erasing and
writing in new rules in his first set, and a third set which tells him how
to edit the second, and on and on until we asymptotically approach sets of
books for which I/O is practically meaningless. Now, in truth, the books of
the mind are not truly leveled like this but probably exist on multiple
levels at once, with different types of information in the brain having
widespread effects on many other types. Even sets of remarkably few neurons
will demonstrate very, very complicated recursions. If each neuron has its
own rules for varying connections based on input and prior connection
strength (much like the rules for a TM,) the fact that it can change it owes
rules perhaps lends itself to the idea of mental non-computability, at least
in today's sense of the word.
Swobe is still wrong, but brains aren't Turing equivalent because the brain
does NOT remain a constant T(n) but instead is composed of innumerable
modular T(x); T(y); T(z); each is constantly changing the T-value of itself
and adjacent virtual machines. Each module has in it some semblance of
UTM-ness allowing it to read others, perhaps owing to a greater mental
structure of which we are not yet aware.
I understand the physicalist's desire to immediately quash all notions of
noncomputability, but this is the same sort of blind partisanship that, if
continued, will prevent us from truly learning how we think. A static TM is
a limited concept. Understanding of the brain will dictate our need to
branch out and explore self-modifying Turturingmachineing Machines...
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the extropy-chat