[ExI] Scientists Discovered ?Mini-Computers? in Human Neurons?and That?s Great News for AI

Henrik Ohrstrom henrik.ohrstrom at gmail.com
Sun Jan 19 13:28:26 UTC 2020


The number of computations in any cell when you calculate the Landauer
energy expenditure needs to include all the housekeeping calculations that
any cell do.
That is a rather large number of computations, almost all done in a rather
sloppy and partially reversible manner.  (Saves energy as they are at least
partially reversible? )
Any synapse should be regarded as a computational hotspot where you have an
output that is both digital and analog as it at the same time can transmit
all kinds of modulations on the downstream use of the signal and also send
a feedback upstream to modulate intensity of signal and both amplify and/or
inhibit transmission of the same or other signals. Also the fact that the
signal pathway is in use change the cell with different kind of more or
less persistent changes both in the synapse and all the way back to gene
expression in the nucleus. All these computations needs to be included in
the energy expenditure for any signal.... Get messy rather quickly.
That's just what's happening inside the cell, any nerve cells that are
highly active will also create a change in their support infrastructure,
glial cells Schwann cells and nearby capillaries.
Even more mess....

A quick check in my library tells me that I have not gotten any new books
on the subject since -97.
Meh, nowadays I just whack the CNS black box with chemicals until it
respond in a somewhat useful manner.
Do no more exact estimates of energy. Need to get updated, again, it never
ends.
And that is a good thing, else it would be booooring.
/Henrik


Den sön 19 jan. 2020 13:46John Clark via extropy-chat <
extropy-chat at lists.extropy.org> skrev:

> On Sat, Jan 18, 2020 at 9:31 PM Stuart LaForge via extropy-chat <
> extropy-chat at lists.extropy.org> wrote:
>
>>
> > Landauer's law can't be responsible for the high energy usage of the
>>> > brain, dendrites are not nearly small enough or fast enough for it
>>> > to be important. At body temperature Landauer says you could erase
>>> > 10^11 bits a second and use only 2.8* 10^-10 watts of energy.  And
>>> > remember signals in the brain only travel at a few hundred meters a
>>> > second. Even our best microprocessors are much too large and much
>>> > too slow for Landauer's law to be important, although in about 20
>>> > years that could change and we'll have to start thinking about reversible
>>> >computing.
>>
>>
>>
>>
>> * > Landauer's law is responsible in that it is defined as the lowest
>> possible energy it would take to erase a bit in a 100% efficient  computer.
>> *
>
>
> Yes Landauer's law will tell you the least amount of energy needed to
> erase one bit of information, but if you want to use more energy than that
> it's OK with Landauer, and the human brain uses trillions of times more
> energy than that lower limit.
>
> *> It is like a thermodynamic limit*
>
>
> It gives a lower limit but not a upper one, you are free to use as much
> energy as you want to erase one bit of information. And the brain uses a
> LOT more.
>
>
>>
>>
>> *> and it costs  biological brains more to erase bits because they are
>> kludgy  naturally-evolved processors that uses selective movement of ions
>> across membranes*
>
>
> Yes, and that kludgy nature is the very reason Landauer's law doesn't
> play an important role in the brain. Nothing in biology is anywhere near to
> running up against Landauer's limit, it's far far too small, Biologists
> can safely ignore it.
>
>
>
>> *> rather than a simple current of electrons to perform calculations.*
>>
>
> That's much less kludgy but even todays microprocessor designers can
> safely ignore Landauer's law and can continue to do so for another decade
> or two because factors that have nothing to do with Landauer produce much
> more heat. When they've conquered those factors then they can worry about
> Landauer.
>
> *> Seems like it should be a testable hypothesis.*
>>
>
> It is. You gave the formula for the minimum amount of energy needed to
> erase one bit of information in your last post, just plug in the numbers
> for body temperature and the formula will give you a ridiculously small
> number, a number that remains ridiculously small even if you multiply it by
> 86 billion (the number of neurons in the human brain) and then multiply it
> again by 15,000 (the maximum number of dendrites any neuron has been
> observed to have). Biologists would do well to spend their time studying
> things other than Landauer, it just isn't important for there subject.
>
> John K Clark
> _______________________________________________
> extropy-chat mailing list
> extropy-chat at lists.extropy.org
> http://lists.extropy.org/mailman/listinfo.cgi/extropy-chat
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20200119/f93b6a20/attachment.htm>


More information about the extropy-chat mailing list