X-Message-Number: 3582
Date: Tue, 3 Jan 1995 00:20:57 -0800
From: John K Clark <>
Subject: SCI.CRYONICS Uploading Yourself
-----BEGIN PGP SIGNED MESSAGE-----
In #3568 (Thomas Donaldson) Wrote:
>It seems to me very unlikely that older people lose their
>memory simply because they've run out of space. In the first
>place, not ALL older people lose their memory.
I think it's safe to say that virtually all old people find
learning more difficult than when they were young. How many
great mathematicians are over 60, or even 50?
> What I was pointing out was that LTP could not alone explain
>long term memory because if so our memory would be a
>whole lot less than it is.
We may not have an enormous amount of long term memory. I've
heard that some medical specialist were distressed to find that
AI programs of just a few megabytes could make pretty good
diagnoses in their field. Considering the years it took them to
acquire this knowledge ,they expected that a much larger program
would be needed.
>Anyone who wants to be suspended but refuses to
>accept present methods may as well forget it.
I agree with you completely but there's no denying techniques
will change radically in the future, perhaps within the lifetime
of some reading this list.
>I'm in favor of speculation, and believe that it can
>lead somewhere [...] these extend from some variation of
>embalming, to ways of reading out our personalities
>and memories into a compyter.
I'll take that as an invitation to speculate because I think
biology is too restrictive and physics places few limits on
what we may become. Landauer, Bennett and Merkle have shown that
with reversible computing the amount of energy needed to make a
calculation can be made arbitrarily small by slowing down the
calculation a little. Even a small reduction in speed can help a
lot in energy saving, the power dissipation ( per unit of time )
falls as the square of the speed. We'll never run out of energy
to perform a calculation but I suppose we could run out of time
if the universe reverses it's expansion and we have a big crunch.
Even today some electronic switches work 100 million times
faster than neural synapses and nanoelectronic switches would be
far faster. If I had good information about a neuron in your
brain I could replace it with an artificial one, the 10,000
other neurons connected to it would see nothing strange; if my
information is good the artificial neuron will act just like the
natural one. When your satisfied the neuron is working properly
and you are still you I get to work on another neuron. After
doing this for 100 billion times your entire brain is now
artificial, now all I have to do is crank up the "clock speed" a
few billion times.
Each artificial neuron would have to give a weighted value to
it's 10,000 synapses , from zero to some large value. I don't
know how many strength levels would be needed but if nature can
do it so can we, there is after all plenty of room to put into each
neuron a powerful nano computer.
The program would be identical to the one operating in your head
at this moment, it would have to be because at first we won't
understand it well enough to change it. We just imitate what
nature does at the neuron level such as "when neuron A and D
sends a signal to E it sends a signal to G and A". I have an
existence proof that the program works ( you) but have no idea
why. All I need to know is how neurons work not how the brain
works. If the programs are the same the only factor left is
speed of the hardware.
To do all this I need to have some understanding about how neurons
operate but I don't need a high level understanding of the
brain, I can just blindly copy from nature. A chip designer may
have no idea how a program running on his chip operates. A
typist may have no understanding about what he writes, some
would say that if nothing else my post proves that point.
There are much better ways to upload than the one I outlined, I
picked the simplest one for debating purposes and some people
like the incremental aspect. It would be much more desirable but
also enormously more difficult to understand how the brain
works not just how neurons work. It could take uploads 100,000
years of vigorous research to figure it out, that's about an
hour in our time.
In discussing uploading I've concentrated on the speedup aspect
for 2 reasons.
1) In the simplest uploading strategy the " blind copy from
nature" one ,a billion fold increase in speed ( and immortality)
was all you could accomplish. To do more you'd have to
understand the logical structure of the mind or make a machine
that could. This is enormously difficult but not impossible.
2) There's no point in talking about things that are
unimaginable, like a mind immensely more powerful than ours not
just faster.
John K Clark
-----BEGIN PGP SIGNATURE-----
Version: 2.6.i
iQCzAgUBLwkFIX03wfSpid95AQFt6QTvQbTH3ESUx0HqN05KZFDetKrmC8iNLYwn
DMDFSpIj4LfX9pyo9zP4OIsomOl43nkFZxHIZ1aEM0OUJ3gIBa1r0RCbydaGcs12
0ZGkscmwoeNHM5DeM0SLDHU/baDUW2Rj5F/Go9529+OJD0VfxR9VkEyqN0Rouqxe
m113Yz+QiOa1e6b0pep43UKkKSDj1ohK1eWQBfk0/Y8xlFYpdPY=
=SQO1
-----END PGP SIGNATURE-----
Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=3582