From: David Cake (dave@difference.com.au)
Date: Mon Jul 08 2002 - 03:09:07 MDT
At 10:16 AM -0600 5/7/02, Ben Goertzel scribbled:
>
>Your disagreement with Edelman is the same as mine: He has this
>stupid idea that connectionist, differential-equations-based
>processes cannot be simulated using discrete digital processes. I
>think he's just obviously wrong.
He is, but for all that he has some very interesting ideas. I
like his ideas about a functionalist definition of consciousness, for
example.
I've been reading his papers for years, and I have just got
used to editing out the bits where he talks about why he isn't a
connectionist.
> I still like his ideas anyway, even though he makes this one
>elementary mistake. I think this mistake is largely borne of his
>ignorance of theoretical computer science and theoretical physics.
>Roger Penrose makes a less ignorant case for the noncomputability of
>brain/mind, though I think he's wrong too...
>
I find Penroses far more irritating, though. I guess its
where they end up - Edelman really wants an analogue computer
substrate, and is just mistaken about how this might be digitally
implemented, but Penroses reasoning about quantum computing comes
uncomfortable close to a scientifically literate explanation for a
version of vitalism to me. He takes all the Chinese room stuff
seriously, and then grasps for something special about the human
brain to justify it, and settle on quantum computing because of its
known computational properties, but his reasoning to get there is
very sloppy.
Regards
David
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:40 MDT