The Singularity has begun!

[bascule]

Okay, not really, but this is sure a big step forward in intelligence research:

 

http://numenta.com/

 

Numenta, a startup company founded by On Intelligence author and Redwood Neuroscience Founder Jeff Hawkins has released both binaries and source code for their general intelligence algorithm, known as Hierarchical Temporal Memory.

 

This software is based on the behavior of hierarchical networks of neocortical columns in the brain, which Hawkins theorizes provide the foundational components of human consciousness.

[Genecks]

Too many theories of neural structure. I'm annoyed by them. I need to see lots and lots of evidence to start believing in any single one. I suppose the release of code is to advertise and lure people into the theory.

[bascule]

Actually, the release of the code is... to foster a community around the code (which he intends to commercialize for solving general intelligence problems, like the kind people are farming off on Amazon's Mechanical Turk). His book "On Intelligence" was to advertise and lure people into the theory.

 

You might want to watch the video. He's presenting before a panel of mostly neuroscientists, and he responds to them at the end:

 

http://video.google.com/videoplay?docid=-2500845581503718756

[foodchain]

It sounds very interesting I have a few questions though and maybe you could clear them up.

 

Is by chance his idea as he would put it for patters in the neocortex or hierarchies of such based more in objective reality of the brain?

 

I ask this because I wonder if the theory itself is a product of the need of modern tools and methods, such as math and how computers work or even what "patterns" the scientist run in there heads for understanding.

 

Second, animal behavior and learning. He touches on learning a lot which I think is rather easy to see. Such is a more historical basis of even psychology going back to Plato. I cant see how you can say its just patterns based on cellular communication or so fourth simply because of the vast difference humans have in that regard with a great deal of life. Simply put to me it seems as if we don’t have any differentiation in that organ and simply put it seems to become as if our brain works simply from bulk of neurons? I don’t know if I follow that one to well. simply it would then seem you could do a count of neurons per specie/organism and find a max or min really of cognitive ability based on neuron count(minus possible compression algorithm I guess), but of course I doubt this to be the case because the brain does have differentiation and related function.

 

For instance, unless routed by specific needs in the overall biology of such the sperm whale has a 20 pound brain roughly, the human brain is a fraction of that in regards to size. The structure though at question, such as the neocortex though, share difference in comparison of the two species.