Is the neocortex only a pattern recognizer?

[Richardn]

Ray Kurzweil said in his How to Create a Mind that to simulate a human brain,both the primitive and the new part must be simulated(this may not be his original words for i read the book in a translated version…this new part refers to the neocortex and the primitive part refers to the basal ganglia and other brain parts).After reading the following contents in the book,I think that his words means:

 --------below is my own interpretation of his words,probably wrong----------

Those creatures that don’t have a neocortex can still make responses to the outside world.For example,a leech which only has tens of thousands of neurons can contract their body when touched and can hunt.They can avoid danger and pursue rewards.For us humans,we also do the same thing.We develop agriculture to harvest crops.We earn money and try to be educated more.These are the behaviours that carry a relatively “hidden” but actually common-in-all-creatures purpose:to avoid death from starvation and to have more chance to reproduce.The only difference between humans and those leeches is that we can avoid highly abstract danger and pursue highly abstract rewards.

 ---------below is my own view,a lot probable to be wrong---------

So why is there the difference?I suppose that it is because we humans have a developed neocortex.It is envolved to efficiently find patterns in the sensory data our body receives.In this way,brains can perform amazing activities like chasing for not so obvious rewards(picking out which actions caused the reward,i.e,predicting),logical thinking(utilizing the patterns in our memories to acheive a new one).This is partly proved by the work you are doing here:using your htm theory based on the structure of the neocortex,you have made great acheivements in anomaly detection.
But I still doubt numenta’s saying:“We believe that understanding how the neocortex works is the fastest path to machine intelligence”.I agree that a developed neocortex makes us different from other creatures,but using this point to prove that the crux of intelligence lies in the neocortex is implying that those creatures don’t have a neocortex are born without intelligence.This is questionable.Chasing rewards and chasing abstract rewards are the same on a larger scale.Something else makes intelligence from 0 to 1 and the neocortex may make it go from 1 to 1000,but the first step is actually more important right?

My very “reckless” conclusion:Neocortex is just a pattern recognizer and a program only consist of htm networks won’t have intelligence.Basal ganglia and limbic system alone makes a worm able to learn.Neocortex alone can’t think.Cooperation of the two makes an AI

 ------questions------

1.How much of the words above is wrong?
2.Why you are so interested in the neocortex?what about other parts of the brain?
3.How is the research of the basal ganglia and the limbic system here going?There are a few posts about this but i can’t get a clear image of the progress and what you have tried on simulating the parts other than the neocortex

 ----------why am i posting all of these?---------

I signed up for this forum yesterday and nearly knows nothing about the htm theory.Before i know this forum, i always believed that the secret of our learning ability,curiosity,fear,etc is in the limbic system,especially in the basal ganglia and have been trying simulating the brain part on a computer.I find this forum full of enthusiastic and knowledgable people devoted to AI research.So i post the thoughts that have been in my head for a long time here trying to get someone willing to criticize or comment on these words.That can’t be more helpful.

[Bitking]

Welcome to the forum; your ideas has been the source of much discussion here.

You may want explore this post:

Discussion about Emotions

Ok I get it I’m the emotions guy… here’s why… your network can’t really learn other than by patterning unless you build a judgement component into it. The emotions serve this purpose for better or worse in the human brain. in this paper- HTMs, Time Locking, and Neural Oscillation (Brain Waves) you may find the reason why we learn. I do not suggest that you try to build a computer with emotions but rather that you consider the value judgement afforded to any decision that emotions bring. An …

This post proposes a model where the “simple” older brain works to train the cortex. It is highly technical and may be a hard read if English is not your first language:

Deep Predictive Learning: A Comprehensive Model of Three Visual Streams

A very recent competing / complementary model of deep predictive coding in the brain. O’Reilly, Randall C., Dean R. Wyatte, and John Rohrlich. "Deep Predictive Learning: A Comprehensive Model of Three Visual Streams." Abstract: How does the neocortex learn and develop the foundations of all our high-level cognitive abilities? We present a comprehensive framework spanning biological, computational, and cognitive levels, with a clear theoretical continuity between levels, providing a coherent…

I personally have been promoting they idea that the older “lizard brain” is the driver of the cortex and behavior for a long time:

Thoughts on multiple pathways for sensory input into the neocortex

It seems to be a thing for connectome/cortex researchers to ignore the processing that goes on in the “lower” brain structures. These were/are sufficient to run lizards and amphibians. There is absolutely no reason to think that they re not performing much of their original functions inside of the functioning brain. The lower brain receives a low-resolution image from the eye area “away” from the fovea and directs the eye to point at interesting areas to form identification of an object with a …

[Paul_Lamb]

@Richardn

Something else makes intelligence from 0 to 1 and the neocortex may make it go from 1 to 1000

I think of the neocortex as an intelligence multiplier, as you implied here in your comment as well. It seems to allow concepts to be taken to unparalleled levels of abstraction, and does it in with (as I understand from the neurosciency folks on the forum) a universal circuit that is repeated across every modality and heirarchical level. The hope is that understanding that basic, repeated circuit well enough to model it in software will allow us to utilize it in other applications (other than brains) to multiply the intelligent behaviors of those systems.

I should point out that I don’t believe modeling the neocortex alone is enough to lead to AGI. I personally don’t even think it is possible to fully implement SMI without simultaneously modeling reinforcement learning. What motivates a system to perform one action over another, or any action at all for that matter without it?

[abshej]

Is it just a pattern recognizer? No.

I think at this point we can say that the neocortex is a memory-prediction organ that not just predicts but also recombines, feeds back and forgets patterns. It is a survival tool necessary for understanding as well as creating abstract patterns. These patterns are sensory and location signals coming from inside and outside the body.
The neocortex is not the control center of the body. It shares function and authority with other organs of the “mind”.
I like to think of the neocortex as something that can breakdown an object(pattern) and distribute its properties to learn them. It can combine patterns and form new abstractions using them as well.
I think the neocortex is a tool of survival that needs supervision to be intelligent.

We are interested in the neocortex because it is the most regular organ in the brain that we can understand with relative ease and with some theory. It is also the organ that has to most functional benefits to us. High level thoughts and tool usage is linked with neocortex. It is also the most recently evolved organ.

Research about Basal Ganglia and other parts is always going on but we have yet to settle on a good working theory of how individual parts function to bring about specific results and how they all work together.

Think of the neocortex as something that allows you to work with immensely complicated tools which other animals can’t.

Also, since we are talking about our current understanding, we need to mention the importance of location sensing in everything that we do. It has become an understanding that human brain does most of the things with a spatial sense.