[extropy-chat] Desirability of Singularity (was Are ancestor simulations immoral?)

[Eliezer S. Yudkowsky]

Damien Sullivan wrote:
> 
>>And the brain is around six orders of magnitude below thermodynamic 
>>efficiency limits for 300 Kelvin...
> 
> Six?  I get 3-5.  Thermo limits of about 1e22 J
> kTln2*bits/s = power
> 1.38e-23*300*.693 * bits/s = 20
> 
> Brain: 1e14 synapses at 1e3 Hz = 1e17 ops/second, with an "op" probably
> being more than a bit/s, so up to 1e19 bits/s.

My computation was based on 1e14 synapses firing an *average* of 20 
spikes per second.  Not every neural circuit is active all the time, and 
1e3 is not realistic even as the rate of a maximally activated neuron, 
except in special cases.

Reality check: a traveling spike involves many ions being released from 
the cell membrane, traveling with the potential grade, then 
painstakingly pumped back in against the potential grade.  It's not 
firing that takes the energy, it's preparation to fire.  If I recall 
correctly, it takes one ATP->ADP reaction per ion pumped against the 
grade.  And I would expect much more than a thousand ions released per 
total synaptic spike.  So the inefficiency relative to the thermodynamic 
limit is surely more than just three orders of magnitude.

-- 
Eliezer S. Yudkowsky                          http://singinst.org/
Research Fellow, Singularity Institute for Artificial Intelligence