From: David Orban (david@davidorban.com)
Date: Sun Sep 02 2007 - 13:11:17 MDT
On 9/2/07, Mike Dougherty <msd001@gmail.com> wrote:
> If each of these machines were a simulated neuron, would the whole be
> approaching the suspected required computational resource of a brain?
Certainly not. Even at the higher range of the estimated number
comprising the botnet of 10 million, we are at four orders of
magnitude less then the number of neurons in the brain (100 billion).
On the other hand, while we are not yet sure of the complete function
of each neuron and their interactions, in my opinion it is safe to
assume that a single PC today is indeed more powerful than an single
neuron.
In terms of topology, and interconnections, it is estimated that there
are 100 trillion synapses connecting the neurons. These are dynamic,
but their flexibility is still likely to be less then the arbitrarily
configurable topology of the botnet.
So: less units, more powerful, and more flexibly connected. In my
opinion, if optimally used, with good software, and confronted with an
environment that enables the software to elaborate a coherent view of
the data supplied, this network would be able surprise us with its
behavior.
--------
David Orban
www.davidorban.com
skype davidorban
sl davidorban
On 9/2/07, Mike Dougherty <msd001@gmail.com> wrote:
> On 9/2/07, David Orban <david@davidorban.com> wrote:
> > There is a nice article about the supercomputing power of botnets in
> > The Washington Post:
> > http://blog.washingtonpost.com/securityfix/2007/08/storm_worm_dwarfs_worlds_top_s_1.html
>
> If each of these machines were a simulated neuron, would the whole be
> approaching the suspected required computational resource of a brain?
> I realize we may not have a way to interact meaningfully with this
> 'thinking' - but that doesn't mean there is not something going on.
> (any more/less than Douglas Adams' idea of the earth being a massive
> supercomputer operated by two mice) We are only recently attempting
> to interface computers with individual neurons, how long before the
> analog of the botnet interfacing with my individual node?
>
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:58 MDT