Upload your mind into a computer by 2050?

[WhapSumi]

CraigD,

 

Please disregard my previous post. I basically speed read your post and missed the fact that you basically already said that...

[alexander]

hey, anyone up for creating 4 dimensional n node dancing tree datastructure to store your brain data and still have some ability to search through it quickly?

[HydrogenBond]

One thing about the human mind that will make it more difficult to transfer is that the human memory is plastic instead of fixed. For example, harddrive memory is fixed and stays put. The living memory does not work exactly this way but is in flux. The living nature of the brain memory creates local and global energy fields about the plastic memories due to the cyclic firing of neurons, as reflected by brainwaves.

 

The potentials within the energy fields, i.e., different memories have different potentials or carry different weight to a person, causes migrations of potential within the global fields, i.e, brain storms. This flux migration of energy can cause the plastic memory to alter within the energy flux. A snap shot today may be different than one in the near future. By then, the memory will contain an element of 20/20 hindsight that came about from the movement of the energy fields in their attempt to lower global potential.

 

A good analogy is like the weather on earth. The sun or sensory systems are evaporating water, which begins to organize itself with other similar memories to form clouds. The storms and rain alter the surface of the earth, i.e., semi-fixed memory. It can wipe out areas and cause plants to grow or die due to rain or drought. Major innovations can also cause hurricanes in the energy field alterring the surface of the earth, i.e., the way we see reality. Newton's telescope created one such hurricane.

 

If one could create plastic computer memory calibrating memory at different potentials or weighs, and then activate them all at the same time, the uneven energy fields will attempt to lower global potential, altering the nature of the plastic memory. If data is constantly inputted at the same time, energy fluxes will mold the dual memory leading to living computers.

[WhapSumi]

If one could create plastic computer memory calibrating memory at different potentials or weighs, and then activate them all at the same time, the uneven energy fields will attempt to lower global potential, altering the nature of the plastic memory. If data is constantly inputted at the same time, energy fluxes will mold the dual memory leading to living computers.

 

This doesn't have to be hardware. This could be done whith software or could be hard coded into the OS for what ever platform you will be importing the contents of your brain to (Lets just hope Microsoft is long gone before then....)

[alexander]

hey Hydrogen, tht is why i said a 4 dimensional dancing tree structure... you can actually emulate brain structures with it :hihi:

[Qfwfq]

No, I haven't lost my mind... it's backed up on tape somewhere!

 

:) :D :) :hyper: :hyper: :hyper: :hyper: :hyper: :hyper: :hyper: :hyper: :hyper: :hyper: :hyper: :hyper: :hyper: :hyper: :hyper: :hyper: :hyper: :hyper: :hyper: :hyper: :hyper: :hyper: :hyper: :hyper:

[alexander]

lol, there we have at least one person who knows how incredibly complex a 4 dimensional dancing tree data structure will be to create... like monts and months of work....

[CraigD]

hey Hydrogen, tht is why i said a 4 dimensional dancing tree structure... you can actually emulate brain structures with it ;)

This is just too intriguing a claim to ignore.

 

OK, Alexander, how does one go about emulating brain structures using a balanced binary tree? (This really deserves its own thread, so let’s start one, or move post to one later)

 

As I understand them (I’ve worked with b-trees for decades, but never one that uses the “dancing” technique to optimize for slow storage access speed), dancing trees are functionally identical to any balanced (no leaf node deeper than any other) b-tree, so any robust b-tree implementation should work well, especially when they use super-fast, disk-in-memory storage devices like EMC drives (which I just happen to have a few idle ones lieing around :thumbs_up )

[Mercedes Benzene]

I just don't see how one could possibly convert the brain's signals into something readable (and interpretable) by a computer.

I would say "No" to uploading your mind into a comp. by 2050.

[HydrogenBond]

The brain has 3-D memory storage capability. This 3-D memory is able to interact in 3-D. The 3-D memory is not implicit of a 3-D geometry of storage but is a type of 3-D logic system of many causes, affects and their integration. Most software is based on logic or 2-D and may not be able to fully simulate 3-D memory.

 

For example, the unified field theory of physics is 3-D, since in integrates all the four forces of nature in a spatial way. It is multiple causes and affects all integrated in 3-D. A computer simulation would only work if all the equations to express this 3-D are already defined. But this is proven to be very elusive. Picture 3-D memories of far more complexity, with no language yet available to translate the hundreds of overlapping causes and affects, all integrated into one in 3-D memory, that itself can interact with other 3-D memories in a way that is 3-D.

 

As a way to visualize 3-D memory, picture it as a ball. With many logic planes of cause and affect intersectng the ball at many different angles, expressing its 3-D volume. When 3-D memory interacts, internal changes within the 3-D memories will impact many overlapping logic planes at the same time. The output to consciousness may be a creative result. It may appear to defy logic only because 3-D organization is not purely logical but has another dimension extra to it.

[alexander]

Ok, craig, how about this as a preview:

 

how do i intend to model brain structure with n-node 4 dimensional tree structure (oh and it would have to hold objects)?

 

simplistic neurobiology:

the brain is a chemicoelecrical organ, it consists of two basic types of cells namely: neurons and glia, each with its own function and with its own set of cells that perform different functions within the 2 classes. A bunch of neurons connected together form a neural network, which is similar in many ways to a circuit board, but with one minor difference, that one neuron can be connected to thousands of other nerons, neural networks put together form neural systems, such as the ones that are responsible for your perseption for example.

how brain works in 3 words or more...:

brains receive signals through nerves arriving from the sensors of the organism, the signal is then sent through the central nervous system, reactions are formulated based upon reflex and learned experiences or ouput given by the neurons.

 

now the tree structure:

modeling neurons in a computer world is nothing less then crazy, imagine a one dimensional tree (aka a node) as a single neuron, now that node contains data or an object that can have actions performed or can perform actions on others, much like a neuron, now a simple network or a 2 dimensional tree (as trees are not limited by the amount of nodes that can be connected to them, not fully flexible n-node trees anyhow...) so you have a neuron and it's neighboring neurons, now a 3 dimensional array, would represent a system of neural networks, where each node is its own network, and the forth dimention is time...

[CraigD]

… modeling neurons in a computer world is nothing less then crazy, imagine a one dimensional tree (aka a node) as a single neuron, now that node contains data or an object that can have actions performed or can perform actions on others, much like a neuron …

You’re describing a Artificial Neural Network. The basic processing element is termed a neurode. This was a tremendously hot Computer Science topic in the mid 1980s, and has produced many valuable commercial applications. The work of Moravec, who I mention in the beginning of this thread, is extensively involved in ANNs.

 

Originally, researchers in ANNs tended to call it “synthetic psychology”. Of late, the term “artificial life” is more popular.

 

I got my introduction to this stuff from Valentino Braitenberg 1984 book “Vehicles: Experiments in Synthetic Psychology”. Since then, there appear to have been a lot of Braitenberg vehicle simulators written, many in Java.

 

The literature on this looks as if its gotten pretty large. Braitenberg’s now at The Max Plank Institute for Biological Cybernetics, which appears to be a great place to be if you’re interested in modeling brain structures in software.

 

Developed as the discipline has become, it appears to be still far away from the goal of uploading a human mind into a computer.

 

PS: In Computer Science, the term “tree” refers to a data structure that is intended to be traversed in some orderly way – it doesn’t have much relationship to the idea of a neural network.

[InfiniteNow]

The brain has 3-D memory storage capability. This 3-D memory is able to interact in 3-D. The 3-D memory is not implicit of a 3-D geometry of storage but is a type of 3-D logic system of many causes, affects and their integration.

Emphasis added by me. You speak in the absolute voice, then you absolutely must support your claims. This has become more common in your posts HB, and you should stop making such unsupported claims immediately. Do not stop making claims, just support those that you do.

 

Support is not personally conceived visual examples, but studies, evidence, and supporting theories which themselves have been tested and stood up to criticism.

 

Provide proof or delete your post.

[CraigD]

The brain has 3-D memory storage capability. This 3-D memory is able to interact in 3-D.

I think I follow what HBond is saying. Microscopic anatomical studies of the human brain do indeed show that its roughly 100,000,000,000 neurons are very richly connected (each to about 10,000 others) in many directions, not constrained to any particular 2-dimensional plane. “3-D storage” is an apt description of this structure

The 3-D memory is not implicit of a 3-D geometry of storage but is a type of 3-D logic system of many causes, affects and their integration. Most software is based on logic or 2-D and may not be able to fully simulate 3-D memory.

I don’t know what “3-D” or “2-D” logic means.

 

Neuroanotomical studies show that neurons function according to complicated dynamics involving the transfer of sodium and potassium ions across their cell membranes. It’s speculated, but not convincingly proven, that this can be modeled as a “summing trigger”, where inputs into a particular neuron’s from many other neurons’ incoming axon terminals “sum” to a numeric value, which, when reached, trigger the neuron to input a variable chemical signal to the multiple neurons to which its axon terminals connect. The synapse connecting neurons have their own complicated dynamic, effecting the overall “logic” of the brain.

 

Geometrically, any finite 1-dimensional array of connected nodes may be mapped to any 2-dimensional array, any 2-D array to a 3-D array, or any n-D array to any (n+1)-D array. (this is a weakened consequence of the diagonalization proof), thus, the connection geometry of a neural network can be considered anything from 1 to (number of nodes -1) –dimensional. What is critical is that nodes can connect to one another without disabling interference. This is accomplished in most electronic logic circuitry (which actually must be 3 dimensional) throught multiple insulating and conduction “chip” layers, or even through 3-dimensional arrangements of separated or insulated wires. In the brain, it is accomplished through a supporting structure of glial cells, and myelin sheaths surrounding the axons of individual neurons.

 

There is no reason to believe that, were such incredibly fine and complicated surgery possible, the brain would not continue to function normally if rearranged into a thin volume resembling a computer chip. The network structure, not the space in which it is arranged, appears to determine the function of the brain.

 

The only credible objection (of which I’m aware) to the proposition that a human brain can be modeled using digital computers is that of Penrose and proponents of his and similar ideas concerning the physics of consciousness. If this position is correct, the only mechanism that will be able to model a human (or likely any animal brain) will be required to have some features essentially identical to the brain.

[InfiniteNow]

Interesting point CraigD. I've never thought of myelin sheaths as adding dimensionality to neuronetworks, and always relegated them purely to functioning to increase speed of signal transmission.

 

I believe that another aspect being left out is the function of the rest of the body. With so much emphasis put on the brain, people often forget that nerve cells populate the entire body and leaving those out would severely limit what is captured by the computer.

[alexander]

PS: In Computer Science, the term “tree” refers to a data structure that is intended to be traversed in some orderly way – it doesn’t have much relationship to the idea of a neural network.

i know what the word tree means in computer science craig, written quite a few trees myself, but sticking to conventional definitions of words in an inventive matter is simply limiting yourself to designing concepts using methods that already exists, not using a unique approach to solve the problem. I do say that there are flaws in my explanation, but well first its not a very complete explanation, however reading more about the brain does reveal to have a certain i should say treeish look to it.... i'm just throwing ideas out, that's cuz i have lots of them (and NSA is spying on me and leaking the information to the manufacturers) ... dont take these quick ideas all personal and shut them down like you already have a working software model of a brain running and its like a big secret and i like just came up with a better way of simulating it or something... i mean i can appologise for throwing a not-so-complete idea up...

and i am not describing an artifficial neural network, my definitions of the brain come from wikipedia (well other sources too, but you will find one that is similar to mine on wiki (used it as a guide)) and if you can capture your brain function for a certain period of time and then come up with an exact model, you should capture all the knowledge that one person has.... there is no way of actually recording all the memory as it is scattered in experiences and reactions of the neurons, there is none nor do i think there will be a direct interface to the brain in the next 43 years and 4 months...

[CraigD]

i know what the word tree means in computer science craig, written quite a few trees myself, but sticking to conventional definitions of words in an inventive matter is simply limiting yourself to designing concepts using methods that already exists, not using a unique approach to solve the problem….

Sorry for my preachy, condescending tone – as my career has scarcely gone a day without an intimate involvement in a particular tree implementation, I occasionally have an inappropriate, proprietary felling about the term. Also, a lot of non-specialists read hypography, and I was concerned they might get some unhelpful ideas about common terminology.

... dont take these quick ideas all personal and shut them down like you already have a working software model of a brain running and its like a big secret and i like just came up with a better way of simulating it or something...

Don’t we both wish?!

 

My intention was just to draw attention to the fact that ideas like yours have a long history, and a lot of active researchers and institutions. Along the way, I learned a good bit myself, as a lot of this research and institutions postdate the time when I actually was working (in the mode of an ill-disciplined but enthusiastic Math major, on quaint hardware, and without notable success) on software models of the brain.

 

Shut you down? That’s the last thing I want to do! I’m hoping that the predictions of folk like Pearson and Moravec are right, and that a way will be found to use future computing resources to run human-mind-equivalent programs, as I’d really like to upload my mind into a computer, before it gets … ehm … “sunsetted” along with its current current meat-host.

 

:) I don’t see how tree managing algorithms are much good for modeling the brain. Fundamentally, they’re about inserting and deleting nodes from a tree, usually in a way that keeps the tree balanced in at least the sense of the terminal (leaf) nodes being of equal depth, in a graph that is distinguished from a general one by having only one path between any 2 nodes. The brain appears to work so well because its “nodes” are connected by many paths, which are managed by a complicated, “squishy” process of physical connection and reconnection. Calling such a graph, a tree, rather than the usual term, a network is, I think, abusing the terms a bit more than is healthy.

 

While new ideas are a good thing – especially in a field that’s not quite lived up to it’s early hopes – I’ve a feeling we’d all benefit from a really deep acquaintance with the current literature. If our minds already were uploaded into computers, we could just crank up our timeslices – unfortunately, they’re not, and all we can do is study and discuss in the same basic way people have been doing it for the last few millennia (though we do have hypography, and gobs of cheap CPU and storage :hihi: ).

there is no way of actually recording all the memory as it is scattered in experiences and reactions of the neurons, there is none nor do i think there will be a direct interface to the brain in the next 43 years and 4 months...

I suspect that, if (and that’s a BIG IF) brain imaging technology with the resolution to see individual neurons can be achieved, we’ll read the whole brain as a single big datum, not as a collection of sequential experiences or ordered knowledge.