[ExI] Universal languages (was: wta-talk Voting Members ...)

[Vladimir Nesov]

On Jan 7, 2008 7:39 AM, Bryan Bishop <kanzure at gmail.com> wrote:
> On Sunday 06 January 2008, Mike Dougherty wrote:
> > destroyed for a single "thought" to occur. I'm not suggesting that we
> > would have any way to interact with such a mind. So maybe your
> > thought experiment could be more limited?
>
> Perhaps. I was also thinking of communicating with contexts in general.
> Like with a rock. Of course, the rock probably wouldn't tell you much,
> as it does not have cognition, and yet it still has properties and it
> is still a statement of the environment that formed it, yes?
>

It may be useful to consider rocks in the same framework: you
communicate with rocks using the same principles. By observing rocks
in their environment you construct their model in your mind, so that
you are able to predict what will happen to them when something
changes or to infer bits of information you don't directly observe.
This communicates information about rocks to you. In opposite
direction, you can create a pattern of rock layout in your mind and
then actually move rocks in the same pattern, thus communicating from
pattern in your mind to physical organization of rocks.

Likewise with all other domains: people understand rules that govern
them, and in cases where they can influence them they can imagine new
patterns that can be communicated to them (created in these domains by
interaction). This way we program computers, engineer new machines,
study materials. Human body replaces old atoms with new ones,
arranging new material in old patterns.

Human intelligence is general enough to understand many things: we can
talk to photons, to galaxies and to other people. But likewise we can
program computers to model some of these things even more precisely,
in cases where unaugmented human mind can't create close enough
analogies within itself.

-- 
Vladimir Nesov                            mailto:robotact at gmail.com