X-Message-Number: 11735
From: Thomas Donaldson <>
Subject: answers to various, with comments for Bob Ettinger at the end
Date: Wed, 12 May 1999 23:12:09 +1000 (EST)

Hi everyone!

Daniel Crevier asks why I insist on parallel computers. My reasons are
simple: any computer that must deal with the real world will have to be
quite fast, faster than any existing or foreseen sequential processor. It
is just not enough to be a Turing machine: the data will come much to
fast for a classical Turing machine, and too fast for a sequential 
machine. As a very simple example, consider vision. I will note that
this is a function which requires either outright parallel computing,
or highly specialized processors which act in parallel while the rest
of the computer plods along sequentially. To operate in the real world
speed on some things is essential.

At one point one person on Cryonet pointed out the merits of quantum
computing here. If we can really get quantum computing going it will be
a good thing; however since quantum computers work on multiple
possibilities simultaneously, they too count as parallel.

As for the complexity of the computer compared to the complexity of the
world, the problem with a virtual person living in a virtual world is
that so far is it is more complex than the world that complexity isn't
exercised at all. As you well know, when presented with no stimuli at
all, the normal human reaction is either to produce illusions or to 
go to sleep. And even the production of illusions does not last forever:
we need data to think on and data to stimulate us constantly. And so
the complex virtual person behaves just like a simple virtual person,
and is no more alive than if that complexity never existed.

You also made some distinctions: a virtual person would have to have
particular circuits (not fully specified) to be conscious. (I gave the
PC example to force you to consider just what was needed). But
consciousness cannot exist in total isolation; it is always consciousness
of SOMETHING. And without anything which matches the complexity of your
brain, you will cease to be conscious at all. (Imagine the effect if
the only thing you could see or do was to look at a blank white wall.
Sure, if someone rescued you after a few hours, you could revive to
full consciousness. But years of blank white wall would kill your
ability to sense anything else, and so kill you).

Put briefly, we cannot separate consciousness from its subject, and
cannot have a complex virtual person living in a simple virtual world.
That is why the complexity of the world becomes so important; and 
also given the speed with which things happen in the real world, why 
parallel computing would be necessary.

And for other commentators:

As for the consciousness of animals, I will point out that neuroscientists 
have considered this issue. It is NOT just a matter of philosophy and
should not be considered as such. And in those terms, many birds and
mammals have brain structures similar enough to our own that we would
really be straining our ideas to argue that they were not conscious.
Whether or not they could communicate what they were conscious of is a
different question and should be seen as such; and some features might
stand out strongly in what they perceived while for us those features were
almost absent. (Rats have a highly developed sense of smell --- their
world must be much more a matter of many different odors than ours).

And for Bob Ettinger: You have several times suggested that brains may
have features which make them impossible to emulate by other kinds of
devices. One quite major difference which neuroscientists have found
is that brains are constantly rewiring themselves on microscopic scales;
that rewiring is part of the acquisition of true long term memory. And
so not even any existing neural net can really emulate the workings of
a brain. However I am not convinced that we could not build a device
which would also have that feature (though it would most certainly work
differently from any existing computer). For that matter, it's not
certain that such features are required for consciousness --- though
we might be able to do some computer tests which might through light
on that issue (a computer behaving as it could rewire itself could
only do so for a short time, but that might be enough).

			Best and long long life to all

				Thomas Donaldson

PS: About the worth of this topic: some cryonicists, as I understand, 
really do think that they will be brought back in computers. We are
discussing this possibility and its plausibility. Those who already
accept that they will NOT be brought back in computers are welcome 
when they choose not to pay attention to this discussion. It's not 
meant for them anyway. 

And yes, I personally am very interested in work to improve our 
suspension methods and ideally prevent all damage to our brains. I'm
happy to give money to any agency which plausibly promises to work on
that problem. But cryonics raises many more questions than the bare
one of survival, though survival remains primary.

Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=11735