[ExI] The Parallel Man

[Eugen Leitl]

On Mon, Oct 10, 2011 at 02:40:20PM -0700, Ben Zaiboc wrote:

> This proposal isn't the same as Moravec's replacement scenario, and there is no disconnection of neurons.
> 
> The idea is to create a simultaneous and parallel information-processing system that 'shadows' what the bio brain is doing, 

This is a classical in vivo incremental uploading scenario.
Notice that the hardware will have considerable bulk and power
footprint, and will utilize numerical methods we do not yet
have.

> adjusting itself and learning until it can do this shadowing perfectly.  
> The reversability of the process lies in the fact that there is no hardware 
> in the brain itself at all (beyond the ion channels or whatever), and no 
> damage is done to the neurons, no disconnection, no removal, nothing like that.  

Sorry, this is bogus. Instead of active grid probe spaced every micron
apart or less and in-situ online processing you're suddenly dealing
with an unphysical load of I/O. I/O is OPS/s, it takes J/s, and the
infrastructure for it is both bulky and invasive.
 
> The 'shutting off' of parts of the bio brain would be via local anaesthesia, 
> transcranial magnetic or electrical stimulation (or should that be depression?), 
> or something similar.  Something that soon wears off, leaving the bit of brain 
> none the worse.  And it would only be used to test if the subjective experience 
> of running in bio or in silico was indistinguishable, and if not, to investigate 
> why not.  Who knows, it might even demonstrate that uploading is actually not 
> possible!  Not that I'd put money on that, of course.

Why do you bother with substituting small areas incrementally?
You realize you can't switch them back on without diverging, and
of course your infrastructure has to speak a lot more than just
electronese. How good is your protonese? Neurotransmittese?

If you want to switch over to shadow infrastructure you do
it instantly, and shut down biology instantly to prevent 
two minds briefly co-existing, and having a chance to diverge.
 
> Even after the subject was happy that the process was complete, there would be no need to do anything to the bio brain (unless you wanted to), and you could happily go for decades with a 'twin-brain', the bio and synthetic parts in perfect harmony.  I'd expect that anyone with this would want to go further, though, and start extending the synthetic part way beyond what a purely bio brain could be capable of, and eventually the whole system would be so much bigger than just the bio component that it would be desirable to ditch it, and go fully synthetic (body as well as brain).  That's when you can start cooking with gas!  Crank the clock speed up, or whatever the equivalent would be, for a start.

Meanwhile, people are dying, and turn to carbon dioxide. Unnecessarily so.
A classical case of misplaced priorities.
 
> I liked the idea of using 3 separate chemical gradients for obtaining a neuron GUID, that's a good one.  I'm also thinking along the lines of the way NMR works, for outputting neuron activity.  Something that can be detected by an external scanning mechanism, rotating through all the ID channels, that is activated only when the neuron fires.

Neurons are not the objects you're looking at.
 
> Crikey, it would have to be fast! 300 billion neurons, isn't it? spike time of a couple of milliseconds?  So.. teraherz region.  Okaaaay, maybe another idea...

Don't spend too much time on this. It's not going to work.

-- 
Eugen* Leitl <a href="http://leitl.org">leitl</a> http://leitl.org
______________________________________________________________
ICBM: 48.07100, 11.36820 http://www.ativel.com http://postbiota.org
8B29F6BE: 099D 78BA 2FD3 B014 B08A  7779 75B0 2443 8B29 F6BE