[ExI] A small step towards brain emulation

Alan Grimes agrimes at speakeasy.net
Tue Nov 20 16:44:57 UTC 2012


Anders Sandberg wrote:

> OK, we seem to be on track when it comes to computing power for big
> neural networks, and the connectomics people are charging ahead (but
> they need to get onto an exponential track of scan-to-neuron conversion:
> right now algorithms are too slow even for the small scan volumes we
> got). It feels like the ball is now in the computational neuroscientist
> court: we better start working on how to turn imaging data into
> something that can run on something like the Compass.

The imaging data will never be useful for anything beyond a point of 
reference.

Because I'm not an uploader, I couldn't care less whether the neural 
simulation is biologically plausible. I only care whether it is 
equivalent or superior, computationally, to the original. I am only 
interested in obtaining an AI. Therefore I'm concerned about getting the 
best AI money can buy. If it would make the AI more powerful, in some 
way, then I'd prefer it look nothing at all like the brain.
Therefore, I'm far more concerned about the three orders of magnitude 
slowdown issue. (which I've been talking about for years....)

-- 
E T F
N H E
D E D

Powers are not rights.




More information about the extropy-chat mailing list