--- In FairfieldLife@yahoogroups.com, "salyavin808" <fintlewoodlewix@...> wrote: > > > > --- In FairfieldLife@yahoogroups.com, "sparaig" <LEnglish5@> wrote: > > > > You are the one getting confused. Rishi, devata and chhandas -observer, > > process of observation and observed. That is all consciousness is. > > ELectrons are conscious, by that definition. > > Then scrap that definition as it confuses you, electrons > are just negatively charged whirly things bound to the > atomic nucleus they surround, there isn't any way they are > aware which is what consciousness means.
You are confusing "aware" with "self-aware". ANything that interacts with other things, is, by the above definition, conscious. > > > ENd of story. > > > > Now, if you want to get all emergent, consider an infinite (or near > > infinite) collection of thingies that interact with each other. By > > definition, that thingie functions as a pattern recognizer on a cosmic > > scale. > > > > A slightly more elaborate end of story. > > It's what I just said with the addition of the idea that electrons > recognise something, I'll have to check the maths but I'm pretty > sure that for all their potential they just sit there holding things > together, the universe then builds what it can. The two things don't > affect each other beyond that. Luckily for us! > Connectionism took off when Hopfield showed that learning algorithms can be described in terms of the same math used to describe physical systems: http://itee.uq.edu.au/~cogs2010/cmc/chapters/Hopfield/ This insight has been extended in many ways. In a sense, you can talk about universal connectionist systems the same way you can Turing Machines: if the math that describes the way a given system works can be recast in terms of a Turing Machine, then that system can perform the same calculations that a Turing Machine can. Likewise, if the components of a physical system can be shown to interact in the same way that the components of an abstract neural network can, then that physical system can perform the same functions as the abstract neural network. In fact, any Turing Machine can be used to model a neural network, and Turing Machines have been created out of Tinkertoys, for example. >From the other direction, a sufficiently complex and properly designed >connectionist network can be used as a Turing Machine. The only question is: >can such networks arise by accident? We already know that they can via >evolutionary forces in biological systems. It is a mathematical certainty that >given enough time/space/energy, random conglomerations of an infinite >collection of interacting parts will include sub-collections that can function >as connectionist systems. The only assertion that I make that perhaps isn't obvious is that the "whole" of the Unified FIeld (whatever "it" is), acts as such a connectionist system. My intuition says that this MUST be the case, though of course, my intuition is often wrong. L
