[ExI] 1 mm^3 of brain

spike at rainier66.com spike at rainier66.com
Sun May 19 16:25:00 UTC 2024



…> On Behalf Of Jason Resch via extropy-chat
Subject: Re: [ExI] 1 mm^3 of brain



On Sat, May 18, 2024, 2:24 PM BillK via extropy-chat <extropy-chat at lists.extropy.org <mailto:extropy-chat at lists.extropy.org> > wrote:

On Wed, 15 May 2024 at 01:26, Darin Sunley <dsunley at gmail.com <mailto:dsunley at gmail.com> > wrote:


This is an impressive article written by the Gemini AI



>…Nice article. However I disagree with it's conclusion that mind uploading would require so much storage. In previous experiments with reducing electron-microscopic imagery to a connection map, saw a roughly million-fold decrease in storage size.


>…Bostrom and Sanders… 


Ja, Sandberg.  But Sanders can be used as a kind of shorthand for Anders.



>…estimated ( https://www.fhi.ox.ac.uk/brain-emulation-roadmap-report.pdf ) that somewhere around 8,000 Terabytes (at 8 bytes per synapse) is more realistic for what a neuronal simulation of the brain would require… Jason


We still haven’t figured out that special magic of a biological brain.  It doesn’t really rely on memory locations analogous to a digital computer, but rather biochemical patterns that exist (somehow) in this neurological tangle between our ears.  Oh that is some marvelous magic, but I don’t think we will get there by trying to simulate synapses individually.


We used finite element analysis to digitally simulate what a structure will do however, and it works well enough.  We model a continuous structure as a collection of rigid cubes with a known flexible interface between the cubes.  I think of it as the atomic world version of a Reimann sum, kinda sorta.  It works, if you make your elements small enough, just as a Reimann sum works if you cut your strips thin enough.


I hafta think there is some acceptable finite-elementish way to model a synapse that won’t require 8 petabytes.  I don’t even like the term petabyte: it sounds expensive and too much like being attacked by the neighbor’s rabid dog.



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20240519/c18b2aff/attachment-0001.htm>

More information about the extropy-chat mailing list