Visual Processing and Vector Calculus

Thanks, I will check it out.
I am thinking about object-oriented programming.

Then by doing the statistic on the interface connections.
We could merge memories together and be imaginative :grinning:

Like put puzzle pieces together and making Lasagna.

What you are describing is similar to the first efforts that computer scientist started with in the 1950s & 1960s. They used tools like LISP to attempt formal reasoning. These attempts were dismal failures and mostly served to show that this โ€œAI thingโ€ was harder than they thought.
See:


and

In the area of neural research, they came at the question of intelligence by studying the architecture of the brain. Certain points in this effort stand out for me because of the profound implications they offer.

One is that a brain is a huge number of individual processing units working together
This is best shown here:

The second is that the data flows through these maps in reasonably well-defined streams

The third is that cortex, the workhorse of this processing is mostly the same in all these areas
with the basic unit of computation being the column.
Possibly the most influential proponent of this concept is Vernon Mountcastle:

The columnar organization of the neocortex

It is a central feature of the work of Numenta -

They work to understand what this processing is. Look to the papers Numenta is offering to see the work they are doing to tease out this basic function.
https://numenta.com/neuroscience-research/research-publications/papers/
That said, the processing that is done in each area is tailored to the related sensory modality and is mostly arranged in 2D organization in each map. As the data progresses through these streams the contents of the maps look very different.
Cortical maps, James A. Bednar & Stuart P. Wilson
https://pdfs.semanticscholar.org/c959/f2f9ad7b44ea938c622876d919ae3cbb90fd.pdf

These various levels do work together to represent semantic content
Ssee this paper to see how this content seems to be grounded or attached to your various sensory streams:
https://www.cell.com/trends/cognitive-sciences/pdf/S1364-6613(13)00122-8.pdf

In the early stages, this is all about things like lines and edges. As you get deep into the process there are structures that have come to be called Grids which represent things like details of physical locations.


Same cortex, but doing very different things with the data.
Researchers are learning a vast number of things about this grid cell level in the internal representation of data - like how it changes when you walk into a room:

I have posted much in this forum about how the subcortical structures interact with the cortex to add emotional color to memories as they are added, and how this is the basis of judgment.

So - getting back to your question about parsing your sensations - I will add some observations about how even these simple questions are not clear cut. How do you feel about the sun? If you live in a war zone the sun is when the fighting starts - a time for fear. If you live in a crime-ridden ghetto - the sun is when the criminal go to bed and it is a good thing. If you live in a cold zone the sun brings warmth and comfort. If you live in a tropical zone the sun brings discomfort and possible life-threatening damage.
I could go on - but you see that there are many possible ways that your stored memories of the sun could be interpreted.

As far as the sun and the clouds the way that these things are coded in the brain seem to be bumps on these grid maps I mentioned above, with the WHAT and WHERE of them distributed over many maps in the temporal lobe. The rules for how the brain manipulates these representations are just now entering intense research. I donโ€™t think anyone really knows how the brain works with these grid maps yet. It is an area that is ready for the right researchers to make important breakthroughs.

3 Likes

Hmmm, I am thinking about a sequence information process.
And this sequence is determined by personal experience :thinking::thinking::thinking:

If it is a beneficial procedure, people will repeat it otherwise abandon it.
What left is so-called reasoning.

Thanks for all those juicy knowledge :yum:

I look forward to seeing your working code.
Best of luck in your endeavors.

1 Like

Yeah, it probably will take few decades.
There are so many things I need to learn :blush:

1 Like