[ExI] Simulating the brain

[Stathis Papaioannou]

On 12 July 2017 at 15:56, Stuart LaForge <avant at sollegro.com> wrote:

> Stathis Papaionnou wrote:
>
> >Calculus works on computer simulations and they are discrete. And if the
> >world really is continuous, it can be simulated on a computer to an
> >arbitrary level of precision. If the 50th decimal place of any physical
> >parameter in your brain is essential to your consciousness, you could not
> >survive, as you would be instantly destroyed by thermal noise.
>
> Even if the brain itself does not have an explicit mechanism to access
> these infinities, if they exist in the brain's environment, they could
> affect brain function. Is your consciousness destroyed by hanging upside
> down? No, of course not. But is your consciousness affected by hanging
> upside down? Probably. You are unlikely to perform as well on an iq test
> while hanging upside down for example. How about if I were to slowly
> adjust your angle relative to gravity until you are at 180 degrees. At
> what point would your mind "change"?
>
> Any arbitrary decimal approximation of the continuum loses an infinite set
> of possible values that are no longer accessible. Moreover, those lost
> values are uncountably infinite so you are losing *amost all* of the
> possible values you had to begin with.
>

Angular displacement of the body will have an effect on neurones, perhaps
by stretching the cell membrane and hence altering the excitability
threshold and the propagation of the action potential. An accurate model of
the brain should therefore take this parameter into account. However, at
some level of resolution the effect will be swamped by noise. So it would
be wasted effort to model angular displacement to 10 decimal places when -
again in order to be accurate - you would have to throw away 5 decimal
places due to the thermal noise inherent in a biological system at body
temperature.


> >Human understanding of irrational numbers does not depend on writing out
> an
> >infinite non-repeating decimal.
>
> Yes. We have the mental capacity to mathematically manipulate infinity and
> discern bona fide truths about infinity without resorting to infinite
> numbers of decimals or infinite memory. On the other hand, I don't think a
> computer has any concept of infinity distinguishable from a stack overflow
> error.


A dog doesn't have much concept of infinity, but its brain is not that that
dissimilar to yours and mine. If we push the point, I don't think any human
can "really" grasp infinity and irrational numbers, even if if they can
manipulate and utilise them as concepts, in the way a computer algebra
system such as Wolfram Alpha can.

>A random number generator could be used for unpredictability.
>
> There isn't any deterministic way of achieving randomness thus random
> numbers generated by computer are pseudorandom and patterns do show up
> upon statistical analysis. Furthermore, irrational behavior need not be
> unpredictable behavior' although it often is.


I don't know of any evidence that a system will behave fundamentally
differently with a truly random as opposed to pseudorandom input, or that
it is possible in general to distinguish between truly random and
pseudorandom.


-- 
Stathis Papaioannou
-- 
Stathis Papaioannou
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20170716/9398a8de/attachment.html>