[extropy-chat] A couple weird questions...

Anders Sandberg asa at nada.kth.se
Fri May 19 22:56:51 UTC 2006


"Hal Finney" wrote:
> That's probably not the right way to look at it.  Each neural firing
> lasts a millisecond or more.  It's a somewhat complex process, involving
> a cascade of ion channels opening, ions rushing in (or out), electrical
> potentials varying, resulting in an explosive climax, not unlike a sneeze.
>
> The bottom line is, from Anders' comment, that a typical neuron spends
> maybe 2% of its time in the firing state.  With 100 billion neurons that
> means that we have 2 billion neurons firing at any one time.

Yes. The data I was referring to (from Cossart, Aronov & Yuste, Attractor
dynamics of network UP states in the neocortex, Nature 423, 15 may 2003)
suggests that each active neuron fires a few times per second during
active states, and that these states appear to be randomly distributed in
time between different populations. So at any given time we probably have
2% in an upstate, firing about 5 times per second, taking up about a
millisecond. So the "duty cycle" for an individual neuron is about
1/10,000. So we have at least a hundred million neurons firing at any
time.


>From a Planck time perspective brain activity is like watching ice ages
come and go (or rather, galaxy clusters to form and dissipate). Even
spikes are slow processes, involving those lazy ions diffusing randomly
through channels and electric fields taking *femtoseconds* to spread
across membranes (28 orders of magnitude difference!).

> Plus even when a neuron is technically not firing, it is still
> doing stuff.  It constantly senses its environment, particularly
> the neurotransmitter concentration.  It may be subject to continual
> small pulses of neurotransmitter that fail to raise its electrical
> potential to the threshold necessary to trigger a firing (like when
> you get that "almost" feeling when you don't quite sneeze).

The Cossart paper shows that the summation of inputs from other neurons
likely is important, since even silent neurons have to detect when they
get a "go" signal to enter an up-state and begin (possibly) firing.

-- 
Anders Sandberg,
Oxford Uehiro Centre for Practical Ethics
Philosophy Faculty of Oxford University





More information about the extropy-chat mailing list