Snowflakes [was Re: Physics - Inertia & Work]

k p Collins kpaulc at [----------]earthlink.net
Thu Jan 8 02:38:45 EST 2004


"Wolf Kirchmeir" <wwolfkir at sympatico.can> wrote in message
news:jbysxveflzcngvpbpna.hr4keq1.pminews at news1.sympatico.ca...
> On Mon, 05 Jan 2004 08:06:25 GMT, k p  Collins wrote:
>
> >Because such asymmetry cross-correlations
> >cannot occur, in an always-occurring way,
> >via simple molecular lattice-building, because,
> >further, if there's asymmetry-cross-correlation,
> >then there must be a mechanism through which
> >geometrical information is communicated from
> >one locus to another.
>
> Not true. All you need is local rules. Play with
> Conway's game of Life and other cellular automata
> to see this. (A google on "cellular automata software"
> will lead you to downloadable freeware.) The
> "mechanism" you think necessary to "communicate
> from one locus to another" is simply the effect of
> the local rules on the adjacent regions, and then on
> the regions adjacent to those, and so on.


Hi Wolf,

Our positions are, at once, not that far apart, and
'at infinite distance' from one another.

They differ only in that, in the view I've discussed,
all  local interactions procede in Deterministic accord
with an overall energy-flow that is the one-way flow of
energy from order to dis-order that is what's =described=
by 2nd Thermo [WDB2T].

It is precisely be-cause 'cellular automata' do not
incorporate any analogous overall dynamics that
they never actually do anything that is Significant,
and verifiably cannot. [The same is True with re-
spect to all 'AI' efforts that I've ever heard of.]

> IMO you also need to meditate on the difference
> between a system that is determined and predictable,
> and one that is determined and unpredictable.

I think about everything that I can think of thinking-about :-]

The only 'difference' is with respect to the information-
processing work entailed.

All available evidence substantiates that physical reality
is Deterministic [all available evidence substantiates that
WDB2T permeates physical reality].

With respect to the analysis of any 'system', the only thing
that varies from analysis to analysis is quantity of information-
processing work necessary to pin things down, and
that quantity of work most often reflects the 'state'
of the data more than anything else.

That is, given a set with 1000 data 'buckets', it's 'hard'
to analyze this set, in a way that enables Prediction, if
980 of the 'buckets' remain empty.

> That difference relates to the difference between
> predictability and explicability.

It's why I've worked at the 'level' of fundamental stuff.
[If I knew of anything more-fundamental than WDB2T,
I'd work at its 'level'.]

And, at that 'level', generalized Predictability is achievable.

Even with respect to the set of all living nervous systems.

So, as above, it comes down to the quantity of work,
and relative-completeness of data.

For instance, although Predicting with respect to the
set of all nervous systems is possible because of what's
in the fundamentals, I'd not venture to 'predict' with
respect to a single nervous system, unless I'd already
accumulated considerable familiarity with respect to
that nervous system's information-processing 'inclinations',
and, then, only with exceeding care - because a 'prediction'
with respect to 'what one will do' tends, strongly, to
result in the occurrence of elevated TD E/I [roughly:
'randomness'] within that nervous system - which tends
to result in that nervous system's converging upon
TD E/I-minimization [roughly: the 'elimination of
randomness'] in ways that 'move away from' the stuff
of the 'prediction'.

In the former case, there's sufficient data. In the latter,
there almost-always isn't.

I hope you'll continue discussing, because I enjoy the
conciseness with which you express arguments.

Cheers, Wolf,

ken [k. p. collins]






More information about the Neur-sci mailing list