Giovanni,

 

Thank you very much, this is an important reference. As you know, I believe 
that the brain implements self-organization, as a side effect of resource 
preservation, by minimizing the length of neuronal connections. From Physics 
alone, not using any input from Neuroscience. The article you found provides 
the first hard experimental evidence that they actually make their connections 
as short as possible, and this happens in many areas of the brain and across 
species as well. I obtained the paper already and will read it.

 

Before I forget, you may want to visit my website SciControls.com, I am working 
on it now and posting new material from time to time. 

 

The other fascinating thing, is that the authors have identified L, the total 
length of all connections, which is exactly the action functional that I have 
proposed years ago. This is the first time that I find such identification. But 
they have no idea about the consequence of doing that, which is that the 
entropy and uncertainty of the information are minimized. 

 

But I have a problem. They say "target points", as if a neuron "knew" 
beforehand where it needs to connect. How can it know that unless it makes a 
connection first? And why do neurons make 10^4 connections each? I thought they 
would make many connections and then select the shortest ones  and discard the 
remaining. And then, the shortest ones happen to be the ones that follow the 
2/3 power law, but this is a consequence rather than a prerequisite. Maybe they 
say something about this in the paper, I haven't read it yet. 

 

My theory says nothing about the causes why the neurons make short connections, 
or make their connections short. I did say something about that myself, but I 
only intended to guess what the possible reasons could be. Any reason that 
results in short connections will minimize the entropy and self-organize the 
information. I always proposed that self-organization and intelligence were 
side effects of the brain making short connections. Why does the brain make 
them short, is another matter, it could be energy, resources, volume, anything. 
Thus, the brain does not know that it is intelligent, and does not intend to 
be. 

 

So you and I agree in full. I don't know much about sleep, but I always thought 
sleep was the time where the brain was at last able to concentrate all its 
resources for shortening its connections or selecting the shortest ones. 

 

Sergio

 

 

 

From: Giovanni Santostasi [mailto:[email protected]] 
Sent: Sunday, August 26, 2012 11:04 PM
To: AGI
Subject: Re: [agi] Uncertainty, causality, entropy, self-organization, and 
Schroedinger's cat.

 

Sergio,
Here some findings about neural connectivity and how the neurons minimize the 
length and volume occupied by the physical connection.

http://www.kurzweilai.net/simple-mathematical-pattern-describes-shape-of-neuron-jungle

I love your approach to AGI (and world modelling in general) to include 
physical constraints and optimization of the brain operation as foundation for 
what the brain does. But maybe besides energy and entropy as fundamental 
organizing factors, one should consider limited resources as volume and 
building materials in the brain as other important physical parameters that 
forces the brain to become an optimization machine.

For example one of the principles that underlies slow wave sleep is the 
renormalization of synapses strength. The brain gets rid of connections that 
are not important (does that were not used often during the day or that didn't 
have a strong signal to noise ratio) during slow wave sleep. The main driving 
force during this process is that the brain needs to be careful in terms of 
resources allocation not necessarily only from an energetic point of view. 

What could minimize energy usage and entropy in a system that has not volume or 
material constraints could be not what works for a system like the brain that 
is necessarily confined by these parameters. 

Giovanni

-------------------------------------------
AGI
Archives: https://www.listbox.com/member/archive/303/=now

RSS Feed: https://www.listbox.com/member/archive/rss/303/18883996-f0d58d57
Modify Your Subscription: https://www.listbox.com/member/? 
<https://www.listbox.com/member/?&;> &

Powered by Listbox: http://www.listbox.com





-------------------------------------------
AGI
Archives: https://www.listbox.com/member/archive/303/=now
RSS Feed: https://www.listbox.com/member/archive/rss/303/17130918-e0ecc803
Modify Your Subscription: https://www.listbox.com/member/? 
<https://www.listbox.com/member/?&;> &
Powered by Listbox: http://www.listbox.com

 


AGI |  <https://www.listbox.com/member/archive/303/=now> Archives  
<https://www.listbox.com/member/archive/rss/303/18883996-f0d58d57> |  
<https://www.listbox.com/member/?&;> Modify Your Subscription

 <http://www.listbox.com> 

 




-------------------------------------------
AGI
Archives: https://www.listbox.com/member/archive/303/=now
RSS Feed: https://www.listbox.com/member/archive/rss/303/21088071-c97d2393
Modify Your Subscription: 
https://www.listbox.com/member/?member_id=21088071&id_secret=21088071-2484a968
Powered by Listbox: http://www.listbox.com

Reply via email to