On 01 Mar 2016, at 04:59, Brent Meeker wrote:
Bruno, what do you think of this paper?
Brent
-------- Forwarded Message --------
The Super-Turing Computational Power of Plastic Recurrent Neural
Networks
http://binds.cs.umass.edu/papers/2014_cabessa.pdf
It is interesting. It might suggest an "easy" solution of the measure
problem of Digital Mechanism.
To get the "super-Turing" power, you still need non recursive patterns
in nature. They suggest it might come from the quantum background, but
obviously, you get it also through the "random oracle" given by the
FPI (First Person Indeterminacy) on all computations (structured by
the modal logics). Such random information has a measure 1 (for the
same reason that we should bet the random "movie" in the iteration of
self-multiplication on all screen images---if you remember the
iterated self-multiplication thought experience). The normal worlds
would come from nature exploiting the (quantum, or computationalist
(below the substitution level)) randomness.
Again, to solve the mind-body problem, such models of computation must
be extracted from the measure imposed by self-reference, but such type
of solution have some plausibility (more than many other more magical
use of Super-Turing (which are Turing + Turing's oracle).
Is it plausibly usable by humans in AI some day? I am not an expert on
this, but I don't see why not.
Bruno
--
You received this message because you are subscribed to the Google
Groups "Everything List" group.
To unsubscribe from this group and stop receiving emails from it,
send an email to [email protected].
To post to this group, send email to [email protected].
Visit this group at https://groups.google.com/group/everything-list.
For more options, visit https://groups.google.com/d/optout.
http://iridia.ulb.ac.be/~marchal/
--
You received this message because you are subscribed to the Google Groups
"Everything List" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
Visit this group at https://groups.google.com/group/everything-list.
For more options, visit https://groups.google.com/d/optout.
