[ExI] Essential Upload Data

[Ben]

On 24/05/2020 14:02, bill w wrote:
> Just to toss in my two cents, one with inflation,
>  What do you think a consciousness is, if not the information that is 
> being duplicated? Your language implies that someone's consciousness, 
> their mind, is a separate thing from the duplicated information.  ben
>
> Well, that's what I think.  I think that consciousness is a dynamic 
> process, which can show up on an EEG,  whereas stored information is a 
> static process (it used to be thought that a memory was a circuit 
> continually running and if it stopped running the memory was lost). 
> Consciousness is the part that accesses the static elements if desired 
> (pulling long term memory into short term memory), along with 
> processing sensory information.  I also would not call consciousness 
> the mind, since most of the mind is unconscious (and static unless 
> called on (?), like accessing the definition of a word).  bill w

I see a contradiction here. You're completely right to say that most of 
a mind is unconscious (so we shouldn't really use the word 
'consciousness' the way we are doing, fair enough), and that the dynamic 
processes are, well, dynamic. But it's not correct to say that dynamic 
processes can't be represented by static data. We do this all the time. 
Again, the example of music comes to mind. So does John Conway's Game of 
Life. The dynamic and self-interacting processes of the game can be 
represented by a simple set of formulae, and additional data about a 
starting state can exactly reproduce the dynamic progression of states 
displayed by the game, when implemented by a suitable computing system.

I see no difference, in principle, between the Game of Life and a mind, 
except for a large difference in complexity. We'll need a lot more 
information to represent the patterns of dynamic interactions, and the 
substrate to run them on will also be (potentially, depending on exactly 
how things are implemented) more complex.

Basically, any information process, no matter how complex, or dynamic or 
self-interactive, can be captured as static data then re-implemented later.

Let's say that it is true thata memory is a circuit continually running, 
and if it stops running the memory is lost.
Let's say that the information in the circuit consists of a specific set 
of spike trains in a loop. If you take the circuit in isolation, and 
break it or stop it, the spike trains are lost and there's no way to 
tell what they were, so restarting the circuit won't make the same 
pattern reappear.

But how did the pattern get there in the first place? There must be some 
configuration of neuronal connections, synaptic weights, and ionic 
concentrations that produced it, looking at the wider brain connected to 
the circuit. So if you can reproduce those, then stopping and restarting 
the whole brain instead of just the circuit in isolation /will/ make the 
same pattern reappear (in fact, our brains depend on this, otherwise 
Marcel Proust would be an obscure nonentity). The same is true of any 
dynamic process in the brain. Or anywhere, for that matter.

You can quote the 'butterfly effect', and say that it won't necessarily 
be exactly the same pattern, but that doesn't matter, for two reasons. 
First, it will be close enough. A small variation of a pattern in a 
circuit won't constitute a totally different memory, as it's the result 
of exactly the same inputs, and second, the butterfly effect won't even 
be a factor in a brain, because it isn't something like a weather system 
with many independent variables, it's a tightly-integrated system with 
many attractors that similar patterns will settle into, just as the 
cells in a tissue do, so that you can provide some of the cues, and the 
cells themselves will provide the rest to settle into a 'mucle tissue' 
pattern or a 'fibrous connective tissue' pattern, etc.

If this weren't true, then people under deep anaesthesia, or in any 
state where their mental processes are interrupted, wouldn't be able to 
resume them (or would display different personalities, be completely 
different people upon resuscitation). All the processes that are 
suspended, are stored in a static form which is then used to restart them.

We don't kill people by anaethetising them, or cooling their brains to 4 
degrees centigrade, therefore dynamic processes can be stored as static 
data.

On a separate note, if your mind is not information, and is not matter, 
then what else could it be? The only things we have to work with (unless 
you think magic is true) are space/time, matter/energy and information. 
Information changing with time is still information (information which 
changes, plus information about how it changes with time. Plus, if 
needed, information about how the changes change, and so on, ad 
infinitum), and we can easily demonstrate that suspending a process in 
time does not irretreivably destroy it.

So, for cryonics to successfully preserve enough information to recreate 
a mind, it looks like we'll need the connectome, and the synaptic 
weights of all the crucial synapses for personality (what these are, we 
don't currently know. I doubt that all of them in the whole brain will 
be needed, but I may be wrong), plus possibly the concentrations of 
certain key ions in certain key places might be desirable too, although 
I suspect that omitting them would just cause a bit of vagueness, and an 
upload becoming conscious would experience some confusion about what has 
happened, much like waking from a dream. But if we have the technology 
to record the weights in billions of synapses, I don't think adding 
information about neuron hillock excitation states would be that hard, 
so waking up as an upload could be a pretty seamless experience. You 
die, then you wake up again, with your train of thought uninterrupted. 
Would that be important? Personally, I'd be happy to forgo it, I think.

-- 
Ben Zaiboc

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20200524/a761e283/attachment.htm>