From: AD.COM - Marcos Guillen (marcos@ad.com)
Date: Tue Sep 16 2003 - 19:34:00 MDT
Hi Ben
>Sorry if my comments seemed too skeptical and not understanding enough of
>the nature of your enterprise. I get that way sometimes, especially when my
>obsession with AI leads me to cut my sleeping hours too short ;-)
It's Ok, Ben, your original comments seem 'fair an balance' enough for me.
You have show a lot of shelf-restrain from your natural bias against, well,
a competing project. After all this is a project that has chosen a very
different approach from yours, and believe me, competing researchers and CEO
of competing companies are not usually as kind as we are being with each
other :-)
Besides, from your point of view, we have probably popped up into existence
from nowhere, so your scepticism is not only unsurprising, but also necessary.
>I certainly agree that having a large-scale distributed platform for neural
>computing will be a very good thing. And I'm glad you agree that, to get
>real value out of this network for AI (or for theoretical neuroscience, for
>that matter), a lot of hard thinking will need to be done regarding the
>specific design of the neural network being run on this exciting new
>platform.
Yes, I agree, a lot of hard thinking is needed, but specially a lot of
testing. That is what justifies building the CCortex Cluster prototype in
the first place.
>I agree that neuroscience, in itself, does not currently provide enough
>guidance to create a highly intelligent, large-scale neural net. What it
>provides is general guidance, and the rest of the architecture of your NN
>will have to be figured out through intuitive combination of information
>from neuroscience, psychology, computer science, with a healthy dose of pure
>guesswork.
Agreed. But as a company we must concentrate on our strengths, software
development and computer infrastructure deployment. We will let other much
better suited groups do the neuroscience research work. Of course we can
only hope they publish it, so that we can read it and incorporate it on our
model.
Incidentally, there are a few millionaires around that are already using
some astonishing money for doing just that, neuroscience basic research.
Jeff Hawkins has been doing that for a few years now, and Paul Allen has
just decided to donate $100 million to help 'decipher' the brain. Hey, I'm
not competing with Paul! I think I let him hire the neuroscientists.
>If you ever feel interested in doing some collaborative thinking on how to
>structure your mega neural network in a useful way, I'm sure that both Peter
>Voss and I would be very willing to toss around some ideas together with
>your team. While we don't have hardware infrastructures on the scale of
>yours, we've been thinking about how to structure quasi-neural computing
>systems for quite some time (although my own AI design Novamente, isn't
>really all that quasi-neural anymore, it did start out that way).
We do welcome the collaboration with other researches. In fact, that was
partially the porpuse of our presentation: To reach groups like yours :-)
Best Regards,
Marcos Guillen
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:42 MDT