[ExI] Vermis ex machina

[Rafal Smigrodzki]

On Sat, Feb 28, 2015 at 11:45 PM, Stuart LaForge <avant at sollegro.com> wrote:
>
>
> I won't argue with your point, since I would be happy with my estimate
> being an upper bound with an error of an order of magnitude. It is
> certainly a much tighter upper bound than the Beckenstein limit which seems
> to give John Clark conniptions. If there are better estimates based on some
> combination of math and empirical observations, I would like to know those
> as well.
>

### Oh, no, I didn't mean to demean your estimate, just added some even
more positive spin on it :)


>
> Although your comment does raise the question of how much redundancy is
> necessary to simulate the human brain. After all computer data is prone to
> bit rot and other forms of data corruption, so for long-term robustness of
> "identity" the redundancy might be unavoidable lest it be satisfactory that
> entropy have its way with uploads.


### My feeling is that a lot in the brain can be vastly simplified. We have
hundreds of millions of cells whose outputs could be probably simulated by
single signals. There are thousands of cells that run your quadriceps but
it's only because you need a lot of wiring to reach each one the millions
of muscle fibers - if you had a hydraulic musculature you would need to run
only a single wire to control a valve to get the same level of mechanical
control. Probably there are many other locations where cell number is
forced not by computational complexity but by the sheer physical size of
resources that have to be controlled.

It's not an accident that whales have large brains.

Also, the brain is made of very low-reliability computational elements,
with MTBF probably measured in hours (a guess on my part, not knowledge).
Running a simulation using more reliable elements would allow for using a
much smaller number of elements. What is the MTBF of a transistor?

Rafał
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20150301/58ce3eca/attachment.html>