I think for them, as for all of us, their biggest danger is hassle from dumb humans.

I don't subscribe to the concept that everything was paradise in the past and/or is paradise in tribal cultures. Many tribal societies are just as bad as western ones in different ways (the taboo system being one prime example, and disease-ridden cockroach-infested villages being another). But here and there we do find pockets of sanity (the Yanomami & the Semai are pretty cool too).

What I want to see is tribal culture unified with modern tech. In the immortal words of the Scal dude, "Sort of like a DSL-compliant Rivendell"  :  )
Best,
AR

Hi dudes, yes you are right Alex, it does seem that the Hadzabi are a rare pocket of tribal sanity. 

Here is another video that to me relates to the idea of I insanity (incongruence) as the situation where people base their lives on something that is unreal. 


http://v7.lscache7.googlevideo.com/vide … mp;key=ck1


Greetings!

Sirhinojo wrote:
Here is another video that to me relates to the idea of I insanity (incongruence) as the situation where people base their lives on something that is unreal. http://v7.lscache7.googlevideo.com/vide … 9C&key=ck1 Greetings!

Videos about insanity or belief in the unreal are not good input, as we don't want to copy (even unconsciously) that sort of behavior. We progress fastest via good input only. Never watch idiots at work.  :  )
Best,
AR

Interesting, looking at it right now.

The grid cells stuff i think it start at around 29:00.

Edit: Another Talk about grid cells, space, etc

http://www.ted.com/talks/lang/en/neil_b … u_are.html

Cheers!

Re: 
http://www.ted.com/talks/lang/en/neil_b … u_are.html

...Umm, can't get this link to work?
AR

Re:
http://www.ted.com/talks/lang/en/neil_b … u_are.html

The distinction between the hippo and entorhinal cortex is not made very clear... I'm aware that some researchers count the ERC as 'part of' the hippo and some count it as part of basal ganglia, but their functions are so different that it's important to know what you're targeting (for example if using NMS or TMS), especially when working with memory. It's also likely that Grid cells are not based on triangles but on hexagons.

What's great is showing graphically how the grid system achieves 3D internal imagery; nicely portrayed. Also it's quite exciting that so many people are getting the concept of 'inner mapping', as this will prepare them for the big surprise when they find out it's imagination running the show  :  )
Best,
AR

Hi dudes,
Re:
the "Next Decisive Years"... (look out, the link is just the first of 3 episodes): http://www.youtube.com/watch?v=_rPH1Abuu9M

Jeff Hawkins (Numenta API for developers - bottom up system/data structure, with some top-down feedback streams) http://www.youtube.com/watch?v=cCdbZqI1r7I


Both these dudes are doing really cool things, and I figured out from this comparison what is the difference and why Hawkins' work is more exciting for me.

Markram's ideal aim seems to be to produce an absolutely accurate 3D 'model' of brain connectivity in a 'typical' brain at a given moment in time. His ideal scenario would be if we could 3D-scan a person's brain with extraordinary accuracy, down to the cellular level, and construct an absolutely accurate model of it. Like an absolutely accurate model of any complex machine, such as a spaceship.

Okay then you got a machine with no software and no 'driver', so you can program in software; getting automatic responses to commands, just like the unconscious mind.

But it isn't, because if we could perfectly scan anyone's connectome, by the time the computer had rendered it, the living connectome would have changed; removed some connections, added others, in response to the experience of being in the real world in real life. And our 'connectome' changes all the time in this way; that's exactly how we get smarter (or more stupid). And our software doesn't do this on its own; it does it in response to real time interaction with the real world of our experience plus its own memory & imagination. What Markram is recreating is a copy of a 'static' brain, and real brains are never static (even in a coma).

So in effect even with a perfect copy, we only have a 'snapshot' of what someone's connectome was like in that particular context of spacetime. By the time a month or two is over, the scanned copy remains the same but the scanned person has a different connectome.

Here's some nice tasty tech that can do a series of such shots capturing plasticity:
http://www.plosone.org/article/info:doi … ne.0041236


I thought some bits were also slightly misleading (One can get 'gamma oscillations' to emerge from any wave generator but that doesn't mean they are being used for the same thing the brain is using them for, or generating them spontaneously). The performance level or 'skill' of any such network at any given task will be in the hands of the programmer.

I don't think Markram is looking for anything else, because there's no need to in the context of this research. This sort of network could do a lot of really great useful things, and probably will.

I'm more interested in the 'not so useful' things that humans do, like learning without being taught, innovation, creativity, self-motivation and autonomous (rather than automatic) response.

Hawkins' work is no less useful for products and services, yet I think there is more there if you're interested in AI for its own (rather than our) sake; because he's working with software (the real 'bottom-up') and copying what real software does. Rather than copying what the brain is like, he's copying the patterns that make it what it's like; the SW that enables emergence, that builds brains.

With my limited knowledge it seems that this approach would be more likely to lead to the design of systems that can not only be interactively flexible (controlled, changed and regulated in response to SW commands, themselves responding to environmental/contextual triggers in the real world), but also SW that can write/upgrade its own SW and develop the equivalent of our value-weighting system and prediction engine. The 'skill' of any such network at any given task will then be in the hands of itself, just as ours is; open to the possibilities plasticity & epigenetics allow.

Markram's system isn't as far as I can tell embodied in a matrix that will allow or necessitate the development of emergence; the first essential for a working mind. Hawkins' appears to have the potential to implement it almost by accident.

...But that's my own 'prediction engine' talking...Maybe of course they are both going to the same place from different starting points, and Hawkins just communicates in a way I find very easy to grasp?

Either way this is all good stuff, not to mention a great neuroscience primer.
Enjoyed,
AR