[frankinstien]

I found some interesting articles on the modeling of cortical columns. While it has been proven that there isn't a fixed set of neurons per column, none the less the structures are similar through out the neocortex.



This is an older model. And this is a much more recent model. There are some ides on how cortical columns enable learning.

I find this kind of research fascinating and inspirational since cortical columns are a similar structure across the brain it implies a similar access point across the various layers of cortical tissue, meaning data can be similar as to how to access it and propagate it, which is how my object-oriented data model came about. Self-similar structure to describe anything, it's almost fractal-like.

[MagnusWootton]

what is that, a colour tv set?

[HS]

This video adds self-assembly and synchronization to the mix, maybe it'll provide some more inspiration. Regarding self-similarity, the researchers were surprised to see dynamics reminiscent of the mammalian cortex developing in the hydra's nervous system, so some essential elements of biological intelligence might be discovered through invariant analyses of different species.

https://www.youtube.com/watch?v=YLssTiSmB28

Although much is known about how the structure of the nervous system develops, it is still unclear how its functional modularity arises. A dream experiment would be to observe the entire development of a nervous system, correlating the emergence of functional units with their associated behaviors. This is possible in the cnidarian Hydra vulgaris, which, after its complete dissociation into individual cells, can reassemble itself back together into a normal animal. We used calcium imaging to monitor the complete neuronal activity of dissociated Hydra as they reaggregated over several days. Initially uncoordinated neuronal activity became synchronized into coactive neuronal ensembles. These local modules then synchronized with others, building larger functional ensembles that eventually extended throughout the entire reaggregate, generating neuronal rhythms similar to those of intact animals. Global synchronization was not due to neurite outgrowth but to strengthening of functional connections between ensembles. We conclude that Hydra’s nervous system achieves its functional reassembly through the hierarchical modularity of neuronal ensembles.

[frankinstien]

I fixed the more recent model link.

[MikeB]

What can you learn from neurons... and for what? As GPT is just guessing algorithms.

There's more to discover on the social side... The minds of creatures that work together vs those that don't. Personality variation to aid the survival of the social group... Minds "aspiring" more when not stressed ("anti-porter x" system in humans). Depression and brain fade from loneliness or anti social personalities

[HS]

What can you learn from neurons... and for what?

For me, the hope with neurons is a unified answer to it all, for the source of awareness, value judgment, and reasoning. If this was discovered and recreated on a sufficiently powerful computing platform, then the resulting AI should be able to experience the world and generate its own understanding. This is the type of AI I'd like to build, and neurons seem like the best way to achieve that. Also, with artificial neurons, I see the potential for scalable self-maintaining designs and medical applications. 

[LOCKSUIT]

You have to look at groups of neurons :). Like car traffic, not just individual cars.

DALL-E/ NUWA is nearly the human brain AI we wanted, all made from testing code and theory, not the brain/skull work with blood all over your hands :). At least, we don't really need to continue looking at the meaty thing now, focus on NUWA.

Yes NUWA uses just a few mechanisms to solve all sorts of problems (image/ video/ texts) thrown at it. That's why the human brain is more patterny then what you even learnt or can see so far!

Things cool down and get less dense after th big bang. Patterns arise after exploratory CHAOS.

The new world made by the next species will be a huge but less dense (so to not become a star/ sun) fractal of metal, like a GPU grid, cold and dark, and 3D, so that they can predict better and expect where and what and when things will be ex. all human food buildings are often grouped near each other, homes are together in another area, and utensils are together in a droor. Things are shaped square because it is predictable and simple, notice every thing in your home is square? Pixels, chairs, TVs, homes, roads, keyboards, beds, boxes, papers, books, sinks. If not square, then sphere. They won't need much light in the future because they will know where things are in their homeworld blindfolded, and can just glide up or down and withstand errors due to reliability (metal, redundancy, cloning) to live forever. This is how I'd want it, to live longer. I will love it soon as my new goal.

Pattern is also life in time in space - lifespan means to persist (pattern) and cloning (is pattern! immortality). If you clone a AGI brain, and resume both, both are the Same thing! There's now 2 of him.

Pattern is also how brains work. They HAVE to recognize old problems to get the solution. It's all matching techniques ex. walk = w a l k = WALK = W A L K = klaw = run = R  U  N = to go very fast = "the vision of this text" (text 2 video)