IUBio

37.7 million artificial neurons

Richard Sargent dsargent at ix.netcom.com
Sun Jan 10 16:41:34 EST 1999



jjj wrote:

> ... is state of the art in computer nn hardware?

nn = neural network?

> Any comments on the article
> by BBC (from New Scientist?)?
>
> jay


    Roboneko (Robot child cat) is a very interesting project with a very large
artificial 'nervous system'.  The large number of artificial neurons in the CAM
('Cellular Automata Machine' brain) is quite impressive (unprecedented
number?).  It's hard to believe an actual simulo-neuron can be created [with its
thousands of dendritic connections], let alone 37.7 million of them.    The
article really didn't go into particulars, such as how the brain would store
memories, create 'burned circuits', or even just how it would have the millions
of reactions that happen progressing  simultaneously.  Instead it dismisses
these as matters that will be solved  by simulated biological evolutionary
steps.
    There are too many questions unanswered and dubious solutions just presented
like the whole biological evolution answer to the different functions inted to
be homologous to living brain functions.
    The nervous system's functions are so complex as to make this project seem
inconceivable to current human capability and knowledge.  Sure, micro-tech is
steadily advancing and circuitry can be made ever smaller but one must consider
the progression to something so incredibly complex as 37.7 * 10^6 neurons each
with x * 10^3 connections (dendrites) which have connections with other
neurons[, the number is incomprehensively huge] as being something that is too
advanced right now.  Instead the computer-biotech field should concentrate on
something involving much fewer base unit cells (I can't think of any examples;
maybe a chicken, a fish, or a grasshopper?).  The project is admittedly
exciting, but it will run into innumerable problems that are unsolvable with the
collective human knowledge of today.  Therefore, Jay, I think this project,
while it is a great statement as regards contemporary microtechnology, is simply
a big waste of money.  That said, I realize that there are many people with
prodigious amounts of money whho are looking for ways to use it.  I'm very
interested in the 'evolution' of this project and if you continue to post
germane Roboneko information with regard to the project's progression, it would
be appreciated.

Thanks again,

Peter L. Sargent




More information about the Neur-sci mailing list

Send comments to us at biosci-help [At] net.bio.net