If the access time is kept minimal thus reducing the number of hard drive read write cycles, it could be possible to use a series of one TeraByte hard drives to shift in and out large sets of neural matter.
The idea is to use hardware Propeller partitions (previously defined) and Neural Containments on the Mac (not yet defined), then load-offload entire partitions, a partition at a time, using Thunderbolt, at potentially 10 Gpbs.
Research has indicated if the hard drive R/W accesses exceed a specific threshold, the drive has a propensity to failure. However, within the confines of this perimeter, full functioning is nominal. What is this parameter?
Those individuals that used large drives for large libraries of iTunes which streamed in constant music led to failures within two years time. Potentially, the drive could be used at these rates and replaced after a year's use.
It is unknown at this time if it's possible to access all of the "supplanted" memory held by the Big Brain's Apple Mac computers. If access is enabled, this would be a boon for the Big Brain. For example, in the Pro version:
Definitions of Supplanted Memory
2.2GHz quad-core Intel Core i7 processor with 6 MB shared L3 cache
4GB 1333MHz DDR3 memory; two SO-DIMM slots support up to 8GB
AMD Radeon HD 6750M graphics processor with 1 GB of GDDR5 memory
Intel HD Graphics 3000 with 384 MB of DDR3 SDRAM
SDXC card slot carrying 16 GB
Optical Drive approx. 650 MB
500 GB ATA HD
USB Sticks 32 GB or
Thunderbolt Exts 2 TB
Total Max 2 TeraByte and 1072 GB and 525 MB
* Not counting additional Macs or Propellers or Hosts
Up until now the Big Brain was retained as an open source development project, where details of development are presented as design and construction continued.
There are obvious constraints of time with this method and explanations of processes do not always clarify functioning, design and development as completely as wished. As ideas could rapidly change and designs modified on the spot, it became a point of question as to which was standardized, while the standard was being developed.
Sometimes a board design was completed and built, only to have it changed 5 minutes later, and negate comments on a previous post. When a board was made to function and then reported successful, it was only met with those who said unfounded claims were being made.
It is also apparent that some posts of information are not being used as intended for the good nature of open source project development. As a result, all open source development is now suspended.
Early on it was stated that open source development was experimental and would be tried - the outcome was unknown. Would it be successful or met with additional challenges? As fate would have it, the additional challenges surfaced in groves.
The project will continue development and postings will continue but not as open source development. This does not indicate the project will or will not be offered open source at the conclusion. Development and designs will take place privately and preliminary wiring schematics and code will generally not appear to avoid any confusion. (test code appears to have upset some who expected finalized code to serve their own project's purpose)
Technology for the Brain will be developed privately, though topical test results can be offered as Brain updates. This change in status may not fare well for those that didn't believe the information previously presented. While there are many people in this world who are liars and cheats, the rest may lose faith as a result. I can say that I have accurately represented the Brain and its development.
What will appear are updates, status reports, general methodology, general ideas, testing results, trends, various other sourced R&D, notes, discussion, sources, links, references, photos and news. I hope this clarifies the current Brain development methodology.
Sometimes a board design was completed and built, only to have it changed 5 minutes later, and negate comments on a previous post
Yeah, that can happen. Perhaps less-frequent, less stream-of-consciousness posts are the answer, rather than retreating behind the wall altogether and arousing even more incredulity.
I'm working really hard on the machine brain for robots. (see brain wiring for Propeller array overview) The latest info is all about adding Macs to the Brain and incorporating those resources. Let's consider the new design overviews and the implications.
Another Big Brain breakthrough!
When the power of the MacPro is added to the Brain, the resources like 4+ Gig of memory are also added. The power of the Propellers is that they now (currently) hold over 100,000 working machine neurons. With the Mac, in the near future, it's possible to put the same structure of these neurons into its resources.
2nd big Brain design breakthrough
The Mac add-on, since going with the more powerful model and updating the Brain design, increases processors from an additional 320 to 720 computers. This is fantastic! I believe it is possible to program these individually for the Brain aggregate, thus adding a truly massive number of more machine neurons. Best modest guess at this point, within the same memory as the programmed Propeller chips, about 720 x 1,000 = 720,000 more neurons. But we could have over a thousand times more memory... making 720,000,000 neurons possible.
A Billion Neurons
When all the totals come in, we'll see a potential one million neurons, leading to the most sophisticated machine brain in the world, in terms of machine neural matter substantive injection. Plus now it's realized there are two other Macs which could add to the Brain. Taking the Mac memory into account, a million neurons could be cloned into one billion. That's going to give the human brain some competition, in terms of neural numbers. This leads to other challenges, i.e. how do you handle a billion of anything? The most likely solution there lies in autonomy.
Coming along the Brain Pipeline
With a financial investment from backers, over 3,000 more computers can be added per board each time with each upgrade or expansion instance, in terms of single board increments, and thus potentially several boards could add on. This truly pumps up the neural matter to over a trillion cloned and exceeds the human brain. What will we do with all this incredible machine Brain power? Anything you want...
I'm very excited about this new approach to a powerful humanoid robot brain, so much that I'm working around the clock on it. I think there is really something in the human spirit that triumphs over adversity. Had we not lost the Brain's initial Windows computer and hard drive, we would not have gained the massively more powerful Brain with Mac computers. Stay tuned as all this development is taking some considerable time.
I think I'm lost. Is the "Big Brain" made up of interconnected Propeller chips or interconnected computers with some Propeller chips linked to it?
I thought 40 Props linked together was intriguing. I've spent lots of time thinking of uses for such a group of interconnected Propellers. One reason for my interest is because I feel like I have a pretty good understanding of how the Propeller works.
I find a group of interconnected computers with some Propellers linked to it a completely different project.
How big a part of the Big Brain are these computers?
I think I'm lost. Is the "Big Brain" made up of interconnected Propeller chips or interconnected computers with some Propeller chips linked to it? I thought 40 Props linked together was intriguing. I've spent lots of time thinking of uses for such a group of interconnected Propellers. One reason for my interest is because I feel like I have a pretty good understanding of how the Propeller works. I find a group of interconnected computers with some Propellers linked to it a completely different project. How big a part of the Big Brain are these computers?
It's the same project, a Big Brain, with included props but always evolving and changing, sometimes on a day to day basis, and the consideration now is to explore use of the mentioned add on Mac computers. It began with the idea of one supporting computer. Then it was realized a leap of neurons could result as part of the system. The Big Brain can use the extra resources from one or more dedicated computers, like hundreds of streaming GPU processors from the native cards, primary quad core CPU and others, and large amounts of memory and devices, for example. One could also see the addition of processor cards if the programming is reasonable. An easy way to look at the prop side is in the initial partitioning which is pure Propeller. Development has included the first, second, and third designed partitions. In the tests performed, partition size was kept variable. The first was over 100 props and then moved down to 60. It was decided to use unlimited expanding Partitions and that multiple partitions built up in increments and groups of 50 that have stable clocking waveforms work best with complete reliability. The recent linked photo shows the wiring of this assembly progressing with the third partition. I hope this gives an overall picture to the Brain status.
Phil: I'm sure there's a balance of posting that can result. Agreed, it's a totally incredulous and fantastic project and some days I can't believe it too!
Youre working on the cutting edge. There comes a time when...
When completed we'll celebrate...maybe a Brain open house day and you can see for yourself. I'm looking at putting together some movies to show it working. There was one posted a while back that showed resulting LED output. Another posted movie helped to inspire at least one new project featured by Parallax. There are some ideas for releasing a smaller version of the Brain for some direct hands on experimenting. It could have a smaller neural package, perhaps 5,000 or 10,000 neurons.
...it's a totally incredulous and fantastic project and some days I can't believe it too!
Are you sure you meant "incredulous"?
From dictionary.reference.com:
in·cred·u·lous [in-krej-uh-luhs]adjective
1. not credulous; disinclined or indisposed to believe; skeptical.
2. indicating or showing unbelief: an incredulous smile.
That may reflect some of your readers' states of mind, but not that of the project itself (unless it's already evolved consciousness: "I think, therefore I am not"?). Did you mean "incredible" instead? Oh, wait: that means "not credible", so probably not. But then you said "fantastic." To which of these definitions (from Google) were you referring?
antic: ludicrously odd; "Hamlet's assumed antic disposition"; "fantastic Halloween costumes"; "a grotesque reflection in the mirror"
extraordinarily good or great ; used especially as intensifiers; "a fantastic trip to the Orient"; "the film was fantastic!"; "a howling success"; "a marvelous collection of rare books"; "had a rattling conversation about politics"; "a tremendous achievement"
fanciful and unrealistic; foolish; "a fantastic idea of his own importance"
existing in fancy only; "fantastic figures with bulbous heads the circumference of a bushel"- Nathaniel Hawthorne
extravagantly fanciful in design, construction, appearance; "Gaudi's fantastic architecture"
Well, anyway, I think all of us are still waiting, after 670 posts, to understand just what is is you're doing -- in sufficient technical detail that it can be replicated.
Well, anyway, I think all of us are still waiting, after 670 posts, to understand just what is is you're doing -- in sufficient technical detail that it can be replicated.
Be patient, Phil. I'm sure the Big Brain will be able to self-replicate soon.
As the Brain is going through its assembly phase, and the buildup of wiring, interfacing and software progresses, many side considerations are opened for thought and contemplation. It is often times through these considerations that new portions of the Brain are added. Other times the consideration doesn't work out for one reason or another and is discarded.
One such consideration is about the topic of a machine Brain DNA and a kind of machine Genome Project. We are already using replicating technology, a simple cloning of neural matter. As the Brain progresses in terms of increasing complexity and definitive constructs and basic algorithms, there is the opportunity to create and define a Machine Brain Genome.
A consulting doctor is working on the greatest Human Genome Project of all time, as millions of Chinese people distributed across China are a part of it, and he has provided some of the most interesting conceptual thoughts towards the Brain project that could be applied to intelligent machines and machine life in my opinion.
As you know, DNA are the building blocks of every life form and the resultant genetic code can tell everything about us. A time is foreseen in the future when machines will have DNA and genetic code.
While we are designing neuronic matter, it seems feasible, at least to consider a layered genetic code. What good is it, you may ask?
http://en.wikipedia.org/wiki/DNA Deoxyribonucleic acid, or DNA, is a nucleic acid that contains the genetic instructions used in the development and functioning of all known living organisms (with the exception of RNA viruses). The main role of DNA molecules is the long-term storage of information. DNA is often compared to a set of blueprints, like a recipe or a code, since it contains the instructions needed to construct other components of cells, such as proteins and RNA molecules. The DNA segments that carry this genetic information are called genes, but other DNA sequences have structural purposes, or are involved in regulating the use of this genetic information.
Thus, it is apparent that machine DNA can contain a set of instructions for the functioning of the Brain and its development. ... all known living organisms ... The Brain is moving toward that definition of a living organism, a machine organism.
http://dictionary.reference.com/browse/organism
–noun
1. a form of life composed of mutually interdependent parts that maintain various vital processes.
2. a form of life considered as an entity; an animal, plant, fungus, protistan, or moneran.
3. any organized body or system conceived of as analogous to a living being: the governmental organism.
Thus, the definition of organism will include a machine life form. It's likely that DNA can be made up of programming software code.
http://cellbiol.com/soft.htm Mac Software for Molecular Biology and Sequence Analysis
Mac Software Download; More Software download sites and Links ... to perform all the basic manipulations of DNA sequences such as restriction mapping, ...
DNA Strider (Mac)
A bit dated but still great full featured sequence manipulation application for the mac. Freeware
Amplify v1.2 (Mac)
This software is for use in designing, analyzing, and simulating experiments involving the polymerase chain reaction (PCR). Given a template sequence and two primers, it will predict the possible outputs of the PCR, taking into account also weak matches to the target sequence. Freeware, thanks to Bill Engels
DNAid v1.8 (Mac)
As the author says, "DNAid is not a heavyweight package which will run your 'just-sequenced gene' through every prediction or analysis algorithm in the known universe." However, it is a very usefull application that allows to perform all the basic manipulations of DNA sequences such as restriction mapping, translation, reverse-complement and others. Freeware, thanks to Fr
Portable battery power remains a challenge after decades of technology development by the battery companies. Typical affordable batteries last on the average of one hour use when powering a processor board and servos(1), depending more or less on power drain and the capacity of the battery. LiPos are new technology and when they run processors, operating time can yield around 7 hours.(2) This is considering just the props or just the computer.(3) Tandem batteries can supply both Brain and computer at the same time to elevate usage time into the 7 hour range. This is preMobility Brain life span, before any servo mechanisms or external sensors are attached.
Let's say we want computational power as the priority and servos are offline during special periods of thought processes. The Brain lasts 7 hours then recharges. So it must maintain a 7 hour cycling time period of maintaining active awake life, while sleeping as it gets a 5 to 12 hour power charge.
When motion mobility is activated, battery drain rises and the Brain has a reduced wake time. If it's designed for a 7-hour work day, now it must nap for recharging.
Another possibility is a circuit to maintain charging during the work day, while working. This probably excludes any major servo drain.
Tests will begin on the new computer in the near future, measuring the computer life span. The Brain Propeller section was already tested for drain, however it is noted that this is only preliminary because the way software is used can determine the power draw. It depends on the number of active cogs, what the code is doing, whether pins are activated, the number of I/O devices, temperature, and chip clocking. It also is determined by the array as a whole, i.e. some props may do A while other props may do B.
A variable condition of power regulation will likely exist. The new computer is rated at 7 hours.(4) The used computer battery lasts fewer hours. The Brain Propeller array is another power variable. It is unlikely these supplies will share direct power, i.e. when the used battery is exhausted, the new battery will not compensate it.
Many of these brain modules and items, particularly the new computer, will serve other functions too. For example, a high end Mac Pro should be used as much as possible to recoup its investment. How this will coexist with Brain life remains to be seen.
1) Robonova
2) Apple
3) Big Brain Propeller Array
4) When on Wireless
You may want to catch up on your Genome reading skills. The very real potential exists for creating Big Brain maps in Genome code rather than circuit schematics. This avoids duplication of Brain Maps.
A machine Genome Map is superior to a schematic because it essentially contains a complete blueprint of the Brain machine and holds more information than a schematic. Machine Genome Maps of DNA Sequences may obsolete the use of schematic diagrams.
The genome is an organism’s complete set of DNA.(1)
http://jgi.doe.gov/education/genomics_1.html What is genomics?
THE HUMAN BODY consists of trillions of cells. Almost all contain an entire genome--the complete set of inherited genetic information encoded in our DNA. When humans reproduce, the parents' sperm and egg DNA combine to contribute a genome's worth of genetic information to the fertilized embryo. That same information is in each of the cells that eventually make up an organism.
Some segments of DNA, called genes or "coding" DNA, contain the chemical recipe that determines particular traits; genetics is the study of the inheritance and function of these genes. Scientists now estimate that humans have about 30,000 genes, located along threadlike, tightly coiled strands of DNA called chromosomes. Genes, however, are only about three percent of human DNA; the rest is "noncoding" DNA. Within these noncoding regions of the genome is the information that determines when and where genes are active--for example, in which cell types and at what stages in the life of an organism. Genomics is the study of the entire set of DNA sequences--both coding and noncoding DNA.
This is the first step of Genetic Big Brain Evolution.
The new computer is on order which will include a genetic genome mapping program that will be applied to the Brain which is a new Machine Organism. Big Brain Genome Mapping is now in effect.
Encoding maps will be derived to specify the Brain's machine DNA. We are sequencing the Brain Machine in terms of the hardware, wiring, interfacing, and options etc. in the first phase I.
What is in Machine DNA? The example below is a simple OUTLINE.
Simple Example of Machine Derived DNA Content Sequence
A(-NUMBER OF NODES
B(-NEURAL MAPPING
C(-TRANSMUTING INTERFACING
D(-COMMUNICATIONS INTERFACING
E(-I/O DESIGNATORS
F(-POSTURALS
G(-SUBMAPPING
H(-EXO
I(-PROPELLER PINNING
J(-DETERMINISTIC ADDRESSING
*In a human, this would be comparable to the color of hair and eyes, structure, gender, skin and other traits
Note that the Big Brain project is defining Machine Genome and Machine DNA.
Genetic Machine Genome Uses
Primary replication, i.e. setting the Big Brain on the Moon and replicating
Transmission of Machine Genome package material for Remote
Big Brain Sex Is the Big Brain Hermaphrodite, Male or Female?
Machine reproduction is an interesting concept because it is highly dependent on the previous topics of Genome and DNA sequences. At some point the machine must become male, female, both or neither.
Compare this to some snails that are both male and female and can do self fertilization. Most species of snails are hermaphrodite, possessing both male and female reproductive organs. It is not usual for a snail to fertilize itself.
Machines are only somewhat Hermaphrodite in nature as they can do self cloning. Self cloning only produces an exact copy of the original and does not produce an intermixing of genome for genetic diversity and evolutionary purposes.
It would make more sense to send one Brain to the Moon and have it replicate itself rather than initially send two brains, one male and one female. This saves expense and weight.
This is the pattern of the initial Big Brain. Later, Big Brains will require sex, Male Big Brains and Female Big Brains - to ensure the mixing of DNA and evolution.
1. not credulous; disinclined or indisposed to believe; skeptical.
2. indicating or showing unbelief: an incredulous smile.
That may reflect some of your readers' states of mind, but not that of the project itself (unless it's already evolved consciousness: "I think, therefore I am not"?). Did you mean "incredible" instead? Oh, wait: that means "not credible", so probably not. But then you said "fantastic." To which of these definitions (from Google) were you referring?
antic: ludicrously odd; "Hamlet's assumed antic disposition"; "fantastic Halloween costumes"; "a grotesque reflection in the mirror"
extraordinarily good or great ; used especially as intensifiers; "a fantastic trip to the Orient"; "the film was fantastic!"; "a howling success"; "a marvelous collection of rare books"; "had a rattling conversation about politics"; "a tremendous achievement"
fanciful and unrealistic; foolish; "a fantastic idea of his own importance"
existing in fancy only; "fantastic figures with bulbous heads the circumference of a bushel"- Nathaniel Hawthorne
extravagantly fanciful in design, construction, appearance; "Gaudi's fantastic architecture"
Well, anyway, I think all of us are still waiting, after 670 posts, to understand just what is is you're doing -- in sufficient technical detail that it can be replicated.
-Phil
Grasshopper: ... “When one eye is fixed on the destination, you have only one eye to search for the way. ....
Big Brain Propeller Waves Different clocks for different blocks
In the human brain we see a variety of brain waves for different purposes. Brain waves result from different functions of the brain. Slow brain waves result from sleep while higher frequency brains waves happen during computational processes. This is analogous to a computer's processor clock. So it is not uncommon for the Big Brain project to have observable brain waves.
The Big Brain can use brain waves to an advantage. In a similar manner to humans, computational processes can have the highest clocks. When isolating parts of the machine Brain, circuits can be clock switched in and out of slower or higher speed, depending on function and purpose.
http://www.brainandhealth.com/Brain-Waves.html
Brain waves are generated by the building blocks of your brain -- the individual cells called neurons. Neurons communicate with each other by electrical changes.We can actually see these electrical changes in the form of brain waves as shown in an EEG (electroencephalogram). Brain waves are measured in cycles per second (Hertz; Hz is the short form). We also talk about the "frequency" of brain wave activity. The lower the number of Hz, the slower the brain activity or the slower the frequency of the activity.Researchers in the 1930's and 40's identified several different types of brain waves. Traditionally, these fall into 4 types:
Delta waves (below 4 hz) occur during sleep
Theta waves (4-7 hz) are associated with sleep, deep relaxation (like hypnotic relaxation), and visualization
Alpha waves (8-13 hz) occur when we are relaxed and calm
Beta waves (13-38 hz) occur when we are actively thinking, problem-solving, etc.
Build a Big Brain Propeller Qualitative EEG Machine Measure the Big Brain's Brain Waves
Human brain waves are measured with Electroencephalography (EEG). The EEG machine will record brain wave activity by the placement of electrode sensors around the human scalp to pick up brain activity.
It's already established that the Big Brain has similar brain waves depending on the Brain's block function and clock frequency. It would be nice to also have a proper EEG machine to measure or simply observe qualitatively the changes in Big Brain wave activity. Now we have such a device.
The Big Brain has its own EEG machine. You can construct this cheap EEG machine using a very simple technique.
Take a portable Amplitude Modulation radio and tune it the low end of the band at a location without any radio stations. Now place it near the Big Brain. Retune the frequency to a gain of strongest signal, i.e. a brain wave signal sound coming from the Big Brain.
As the DIY EEG is moved around the Brain, different sound will result from different clock frequency. Now you can hear and study brain waves emanating from the Big Brain. Clever interpretation of this sound will lead one to recognize what activity is occurring in the Big Brain.
Brain activity can vary depending on locations inside the Brain. Moving the EEG to different parts of the Brain will pick up this brain wave activity. A more sophisticated model of the EEG radio could quantitatively record Big Brain Waves and their locations. Radio, RFI probes placed around the Big Brain can pick up this activity and produce a 3D model of the Big Brain's activity.
In a Big Brain circuit with hundreds of Propeller processors, it may be challenging to quickly diagnose if one chip loses its functioning. Now we've never had any Propeller chip fail and we don't expect any to fail in the future, but there is always the possibility of a wire coming loose and shutting down a prop's functioning. Finding that one prop could be like looking for a needle in a haystack.
By using the Qualitative Big Brain EEG Machine (placed at a more remote distance) and a single probe (on the end of a shielded cable) to quickly scan each Propeller chip, a determination is quickly made as to the operating status of the chip.
A program in SPIN to generate some calculations or a simple loop (REPEAT) would be effective wave generator code for testing Propeller chips inside the Big Brain.
Perhaps a neural injection program could also be adapted to specific Propellers in blocks by injecting neurons instructed to fire at a specific speed.
Do you have an example of a working neuron? Something that can solve the XOR problem would be nice.
Background An artificial neuron is a device with many inputs and one output. The neuron has two modes of operation; the training mode and the using mode. In the training mode, the neuron can be trained to fire (or not), for particular input patterns. In the using mode, when a taught input pattern is detected at the input, its associated output becomes the current output. If the input pattern does not belong in the taught list of input patterns, the firing rule is used to determine whether to fire or not. http://www.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/cs11/report.html
Types of neurons - By connection: There are three classes of neurons: afferent neurons, efferent neurons, and interneurons. Afferent neurons carry information from tissues and organs into the central nervous system. Efferent neurons transport signals from the central nervous system to the effector cells. Interneurons connect neurons within the central nervous system. By function: Sensory neurons carry signals from sense organs to the spinal cord and brain. Relay neurons carry messages from one part of the CNS. Motor neurons carry signals from the CNS to muscles, motor neurons are connected to the relay neurons. The signal passes between the neurons via synapses. Synapses are microscopic voids between cells where chemicals are released from the axon terminal of one cell to specialized chemical receptors on the dendrite of the receiving cell. http://simple.wikipedia.org/wiki/Neuron
At this early machine Brain stage, the fabricated machine test neuron is in a Using Mode with a specific firing pattern that can be varied by programming. The "taught input pattern" is usually a test constant of periodicity, simply made into output on pins. To indicate the test firing and the functioning of the firing in each Propeller chip, fire is indicated by output states. These outputs are seen in some of the early photos in on or off states. The example of the working machine neuron is shown in the illustration below. For the Big Brain, the INIT Neuron (a type of machine neuron for testing) is created. This holds the inputs constant as seen to the left in the diagram and gives several things. One, it produces a single neural matter electronic cell that can be propagated to the array. Two, it gives a periodicity of firing feedback from the Propellers. Three, it indicates the INIT Neuron is firing or not firing. INIT Neurons have four states: constant Use Mode, Fire, Negate Fire, Periodicity.
Yes, that's what I was after. I know something about ANN's and actually supervised a Hull University student's MSc project involving one to detect hand-gestures used for sign-language with a Dataglove when I worked for BAe, MAD at Brough. That was BAe, Miltary Aircraft Division. She used an ANN written in C running on a PC.
Phil & Leon: I doubt the simple test neuron is what you're after. I think both of you are far more knowledgeable and qualified to write a full neuron program - if you want to write it, each neuron would have SPIN code and need to be ultra tiny. Right now I'm trying to do the install of BST on the OSX Mac, so I can get back to Propeller programming, but it comes up with two usb drivers (which one to use?) and then the questions about 32 or 64 bit versions (which one to use?).
@Humanoido, at least give the community the opportunity to view the code. The reader can decide if the "simple neuron test" fits their needs. I'm very curious what a neuron test looks like.
@Humanoido, at least give the community the opportunity to view the code. The reader can decide if the "simple neuron test" fits their needs. I'm very curious what a neuron test looks like.
I didn't post the reason for not posting the code. You may have missed some important posts. We had an unfortunate incident with Windows/hard drive and it looks like 99% of the files were obliterated. The drive had an attempted restore by a professional hard drive restoration company and in the hope of finding working files I continue to sift through so called "restored" files that don't load because they are still irreparably damaged. This is a massive undertaking as the drive has half a million files with changed file names and changed extensions. Don't worry though, if the INIT-Neuron code is not found, a project will begin for new code development. In the mean time, post #684 has a really good description, diagram and mini-tutorial. It's possible I can provide some added text information in the future about how the inputs are soft wired on the test.
Thanks! When I get back today, I'll look at it in more detail.
There's a link with 130 managed projects that look very interesting found here at the root: http://forge.scilab.org/
Partial Sample list below
accsum, Accurate algorithms for sums
Automatic Control Toolbox, Stabilization and tracking toolbox for linear systems (still in development).
aerospace-toolbox, A toolbox for aerospace simulation and design.
ANN toolbox, Toolbox for artificial neural networks
Apifun, An API for argument checking in macros.
Assert, A collection of check functions to support numerical unit tests
autorun, autorun for Scilab on Windows
BackDoor
binary-patch, Binary patching (Gsoc 2011)
Metanet and Boost.Graph
Buffer block, Some demonstration blocks for FFT in Xcos
CelestLab , CNES Space mechanics toolbox for mission analysis
Comments
If the access time is kept minimal thus reducing the number of hard drive read write cycles, it could be possible to use a series of one TeraByte hard drives to shift in and out large sets of neural matter.
The idea is to use hardware Propeller partitions (previously defined) and Neural Containments on the Mac (not yet defined), then load-offload entire partitions, a partition at a time, using Thunderbolt, at potentially 10 Gpbs.
Research has indicated if the hard drive R/W accesses exceed a specific threshold, the drive has a propensity to failure. However, within the confines of this perimeter, full functioning is nominal. What is this parameter?
Those individuals that used large drives for large libraries of iTunes which streamed in constant music led to failures within two years time. Potentially, the drive could be used at these rates and replaced after a year's use.
It is unknown at this time if it's possible to access all of the "supplanted" memory held by the Big Brain's Apple Mac computers. If access is enabled, this would be a boon for the Big Brain. For example, in the Pro version:
Definitions of Supplanted Memory
2.2GHz quad-core Intel Core i7 processor with 6 MB shared L3 cache
4GB 1333MHz DDR3 memory; two SO-DIMM slots support up to 8GB
AMD Radeon HD 6750M graphics processor with 1 GB of GDDR5 memory
Intel HD Graphics 3000 with 384 MB of DDR3 SDRAM
SDXC card slot carrying 16 GB
Optical Drive approx. 650 MB
500 GB ATA HD
USB Sticks 32 GB or
Thunderbolt Exts 2 TB
Total Max 2 TeraByte and 1072 GB and 525 MB
* Not counting additional Macs or Propellers or Hosts
The following links are for development reference when programming in Brain/Mac XCode:
Developer
Mac Developer Program
Technologies
Resources
Support
Member Center
Developer Tools
New in XCode 4
Features
iOS
MAC OSX
Safari
Download XCode
XCode Preview at WWDC 2010
Single Window
http://developer.apple.com/technologies/tools/whats-new.html
Apple LLVM Compiler 2.0
New Debugger
Interface Builder is Built-in
Fix-it and Live Issues
Instruments for Xcode 4
Assistant
Version Editor
Up until now the Big Brain was retained as an open source development project, where details of development are presented as design and construction continued.
There are obvious constraints of time with this method and explanations of processes do not always clarify functioning, design and development as completely as wished. As ideas could rapidly change and designs modified on the spot, it became a point of question as to which was standardized, while the standard was being developed.
Sometimes a board design was completed and built, only to have it changed 5 minutes later, and negate comments on a previous post. When a board was made to function and then reported successful, it was only met with those who said unfounded claims were being made.
It is also apparent that some posts of information are not being used as intended for the good nature of open source project development. As a result, all open source development is now suspended.
Early on it was stated that open source development was experimental and would be tried - the outcome was unknown. Would it be successful or met with additional challenges? As fate would have it, the additional challenges surfaced in groves.
The project will continue development and postings will continue but not as open source development. This does not indicate the project will or will not be offered open source at the conclusion. Development and designs will take place privately and preliminary wiring schematics and code will generally not appear to avoid any confusion. (test code appears to have upset some who expected finalized code to serve their own project's purpose)
Technology for the Brain will be developed privately, though topical test results can be offered as Brain updates. This change in status may not fare well for those that didn't believe the information previously presented. While there are many people in this world who are liars and cheats, the rest may lose faith as a result. I can say that I have accurately represented the Brain and its development.
What will appear are updates, status reports, general methodology, general ideas, testing results, trends, various other sourced R&D, notes, discussion, sources, links, references, photos and news. I hope this clarifies the current Brain development methodology.
-Phil
I think I'm lost. Is the "Big Brain" made up of interconnected Propeller chips or interconnected computers with some Propeller chips linked to it?
I thought 40 Props linked together was intriguing. I've spent lots of time thinking of uses for such a group of interconnected Propellers. One reason for my interest is because I feel like I have a pretty good understanding of how the Propeller works.
I find a group of interconnected computers with some Propellers linked to it a completely different project.
How big a part of the Big Brain are these computers?
When your work is analyzed and critiqued, it feels like a personal attach. I can assure you it’s not personal.
Props and computers
It's the same project, a Big Brain, with included props but always evolving and changing, sometimes on a day to day basis, and the consideration now is to explore use of the mentioned add on Mac computers. It began with the idea of one supporting computer. Then it was realized a leap of neurons could result as part of the system. The Big Brain can use the extra resources from one or more dedicated computers, like hundreds of streaming GPU processors from the native cards, primary quad core CPU and others, and large amounts of memory and devices, for example. One could also see the addition of processor cards if the programming is reasonable. An easy way to look at the prop side is in the initial partitioning which is pure Propeller. Development has included the first, second, and third designed partitions. In the tests performed, partition size was kept variable. The first was over 100 props and then moved down to 60. It was decided to use unlimited expanding Partitions and that multiple partitions built up in increments and groups of 50 that have stable clocking waveforms work best with complete reliability. The recent linked photo shows the wiring of this assembly progressing with the third partition. I hope this gives an overall picture to the Brain status.
1. not credulous; disinclined or indisposed to believe; skeptical.
2. indicating or showing unbelief: an incredulous smile.
That may reflect some of your readers' states of mind, but not that of the project itself (unless it's already evolved consciousness: "I think, therefore I am not"?). Did you mean "incredible" instead? Oh, wait: that means "not credible", so probably not. But then you said "fantastic." To which of these definitions (from Google) were you referring?
- antic: ludicrously odd; "Hamlet's assumed antic disposition"; "fantastic Halloween costumes"; "a grotesque reflection in the mirror"
- extraordinarily good or great ; used especially as intensifiers; "a fantastic trip to the Orient"; "the film was fantastic!"; "a howling success"; "a marvelous collection of rare books"; "had a rattling conversation about politics"; "a tremendous achievement"
- fanciful and unrealistic; foolish; "a fantastic idea of his own importance"
- existing in fancy only; "fantastic figures with bulbous heads the circumference of a bushel"- Nathaniel Hawthorne
- extravagantly fanciful in design, construction, appearance; "Gaudi's fantastic architecture"
Well, anyway, I think all of us are still waiting, after 670 posts, to understand just what is is you're doing -- in sufficient technical detail that it can be replicated.-Phil
Example of Mage
http://kinemage.biochem.duke.edu/kinemage/magepage.php
As the Brain is going through its assembly phase, and the buildup of wiring, interfacing and software progresses, many side considerations are opened for thought and contemplation. It is often times through these considerations that new portions of the Brain are added. Other times the consideration doesn't work out for one reason or another and is discarded.
One such consideration is about the topic of a machine Brain DNA and a kind of machine Genome Project. We are already using replicating technology, a simple cloning of neural matter. As the Brain progresses in terms of increasing complexity and definitive constructs and basic algorithms, there is the opportunity to create and define a Machine Brain Genome.
A consulting doctor is working on the greatest Human Genome Project of all time, as millions of Chinese people distributed across China are a part of it, and he has provided some of the most interesting conceptual thoughts towards the Brain project that could be applied to intelligent machines and machine life in my opinion.
As you know, DNA are the building blocks of every life form and the resultant genetic code can tell everything about us. A time is foreseen in the future when machines will have DNA and genetic code.
While we are designing neuronic matter, it seems feasible, at least to consider a layered genetic code. What good is it, you may ask?
http://en.wikipedia.org/wiki/DNA
Deoxyribonucleic acid, or DNA, is a nucleic acid that contains the genetic instructions used in the development and functioning of all known living organisms (with the exception of RNA viruses). The main role of DNA molecules is the long-term storage of information. DNA is often compared to a set of blueprints, like a recipe or a code, since it contains the instructions needed to construct other components of cells, such as proteins and RNA molecules. The DNA segments that carry this genetic information are called genes, but other DNA sequences have structural purposes, or are involved in regulating the use of this genetic information.
Thus, it is apparent that machine DNA can contain a set of instructions for the functioning of the Brain and its development. ... all known living organisms ... The Brain is moving toward that definition of a living organism, a machine organism.
http://dictionary.reference.com/browse/organism
–noun
1. a form of life composed of mutually interdependent parts that maintain various vital processes.
2. a form of life considered as an entity; an animal, plant, fungus, protistan, or moneran.
3. any organized body or system conceived of as analogous to a living being: the governmental organism.
Thus, the definition of organism will include a machine life form. It's likely that DNA can be made up of programming software code.
http://cellbiol.com/soft.htm
Mac Software for Molecular Biology and Sequence Analysis
Mac Software Download; More Software download sites and Links ... to perform all the basic manipulations of DNA sequences such as restriction mapping, ...
DNA Strider (Mac)
A bit dated but still great full featured sequence manipulation application for the mac. Freeware
Amplify v1.2 (Mac)
This software is for use in designing, analyzing, and simulating experiments involving the polymerase chain reaction (PCR). Given a template sequence and two primers, it will predict the possible outputs of the PCR, taking into account also weak matches to the target sequence. Freeware, thanks to Bill Engels
DNAid v1.8 (Mac)
As the author says, "DNAid is not a heavyweight package which will run your 'just-sequenced gene' through every prediction or analysis algorithm in the known universe." However, it is a very usefull application that allows to perform all the basic manipulations of DNA sequences such as restriction mapping, translation, reverse-complement and others. Freeware, thanks to Fr
Portable battery power remains a challenge after decades of technology development by the battery companies. Typical affordable batteries last on the average of one hour use when powering a processor board and servos(1), depending more or less on power drain and the capacity of the battery. LiPos are new technology and when they run processors, operating time can yield around 7 hours.(2) This is considering just the props or just the computer.(3) Tandem batteries can supply both Brain and computer at the same time to elevate usage time into the 7 hour range. This is preMobility Brain life span, before any servo mechanisms or external sensors are attached.
Let's say we want computational power as the priority and servos are offline during special periods of thought processes. The Brain lasts 7 hours then recharges. So it must maintain a 7 hour cycling time period of maintaining active awake life, while sleeping as it gets a 5 to 12 hour power charge.
When motion mobility is activated, battery drain rises and the Brain has a reduced wake time. If it's designed for a 7-hour work day, now it must nap for recharging.
Another possibility is a circuit to maintain charging during the work day, while working. This probably excludes any major servo drain.
Tests will begin on the new computer in the near future, measuring the computer life span. The Brain Propeller section was already tested for drain, however it is noted that this is only preliminary because the way software is used can determine the power draw. It depends on the number of active cogs, what the code is doing, whether pins are activated, the number of I/O devices, temperature, and chip clocking. It also is determined by the array as a whole, i.e. some props may do A while other props may do B.
A variable condition of power regulation will likely exist. The new computer is rated at 7 hours.(4) The used computer battery lasts fewer hours. The Brain Propeller array is another power variable. It is unlikely these supplies will share direct power, i.e. when the used battery is exhausted, the new battery will not compensate it.
Many of these brain modules and items, particularly the new computer, will serve other functions too. For example, a high end Mac Pro should be used as much as possible to recoup its investment. How this will coexist with Brain life remains to be seen.
1) Robonova
2) Apple
3) Big Brain Propeller Array
4) When on Wireless
You may want to catch up on your Genome reading skills. The very real potential exists for creating Big Brain maps in Genome code rather than circuit schematics. This avoids duplication of Brain Maps.
A machine Genome Map is superior to a schematic because it essentially contains a complete blueprint of the Brain machine and holds more information than a schematic. Machine Genome Maps of DNA Sequences may obsolete the use of schematic diagrams.
The genome is an organism’s complete set of DNA.(1)
1 http://www.ornl.gov/sci/techresources/Human_Genome/project/info.shtml
The DNA Sequencing Process
http://www.ornl.gov/sci/techresources/Human_Genome/graphics/DNASeq.Process.pdf
http://jgi.doe.gov/education/genomics_1.html
What is genomics?
THE HUMAN BODY consists of trillions of cells. Almost all contain an entire genome--the complete set of inherited genetic information encoded in our DNA. When humans reproduce, the parents' sperm and egg DNA combine to contribute a genome's worth of genetic information to the fertilized embryo. That same information is in each of the cells that eventually make up an organism.
Some segments of DNA, called genes or "coding" DNA, contain the chemical recipe that determines particular traits; genetics is the study of the inheritance and function of these genes. Scientists now estimate that humans have about 30,000 genes, located along threadlike, tightly coiled strands of DNA called chromosomes. Genes, however, are only about three percent of human DNA; the rest is "noncoding" DNA. Within these noncoding regions of the genome is the information that determines when and where genes are active--for example, in which cell types and at what stages in the life of an organism. Genomics is the study of the entire set of DNA sequences--both coding and noncoding DNA.
Big Brain Parallax Machine Genome Begins
This is the first step of Genetic Big Brain Evolution.
The new computer is on order which will include a genetic genome mapping program that will be applied to the Brain which is a new Machine Organism. Big Brain Genome Mapping is now in effect.
Encoding maps will be derived to specify the Brain's machine DNA. We are sequencing the Brain Machine in terms of the hardware, wiring, interfacing, and options etc. in the first phase I.
What is in Machine DNA? The example below is a simple OUTLINE.
Simple Example of Machine Derived DNA Content Sequence
A(-NUMBER OF NODES
B(-NEURAL MAPPING
C(-TRANSMUTING INTERFACING
D(-COMMUNICATIONS INTERFACING
E(-I/O DESIGNATORS
F(-POSTURALS
G(-SUBMAPPING
H(-EXO
I(-PROPELLER PINNING
J(-DETERMINISTIC ADDRESSING
*In a human, this would be comparable to the color of hair and eyes, structure, gender, skin and other traits
Note that the Big Brain project is defining Machine Genome and Machine DNA.
Genetic Machine Genome Uses
Is the Big Brain Hermaphrodite, Male or Female?
Machine reproduction is an interesting concept because it is highly dependent on the previous topics of Genome and DNA sequences. At some point the machine must become male, female, both or neither.
Compare this to some snails that are both male and female and can do self fertilization. Most species of snails are hermaphrodite, possessing both male and female reproductive organs. It is not usual for a snail to fertilize itself.
http://wiki.answers.com/Q/How_do_you_know_if_a_snail_is_male_or_female
Machines are only somewhat Hermaphrodite in nature as they can do self cloning. Self cloning only produces an exact copy of the original and does not produce an intermixing of genome for genetic diversity and evolutionary purposes.
It would make more sense to send one Brain to the Moon and have it replicate itself rather than initially send two brains, one male and one female. This saves expense and weight.
This is the pattern of the initial Big Brain. Later, Big Brains will require sex, Male Big Brains and Female Big Brains - to ensure the mixing of DNA and evolution.
http://wiki.answers.com/Q/Why_are_snails_both_male_and_female
"There is a difference in reproducing and creating genetic diversity. Genetic diversity is achieved when two snails mate together."
Grasshopper: ... “When one eye is fixed on the destination, you have only one eye to search for the way. ....
Different clocks for different blocks
In the human brain we see a variety of brain waves for different purposes. Brain waves result from different functions of the brain. Slow brain waves result from sleep while higher frequency brains waves happen during computational processes. This is analogous to a computer's processor clock. So it is not uncommon for the Big Brain project to have observable brain waves.
The Big Brain can use brain waves to an advantage. In a similar manner to humans, computational processes can have the highest clocks. When isolating parts of the machine Brain, circuits can be clock switched in and out of slower or higher speed, depending on function and purpose.
http://www.brainandhealth.com/Brain-Waves.html
Brain waves are generated by the building blocks of your brain -- the individual cells called neurons. Neurons communicate with each other by electrical changes.We can actually see these electrical changes in the form of brain waves as shown in an EEG (electroencephalogram). Brain waves are measured in cycles per second (Hertz; Hz is the short form). We also talk about the "frequency" of brain wave activity. The lower the number of Hz, the slower the brain activity or the slower the frequency of the activity.Researchers in the 1930's and 40's identified several different types of brain waves. Traditionally, these fall into 4 types:
http://blog.makezine.com/archive/2008/11/the-brain-machine.html
Measure the Big Brain's Brain Waves
Human brain waves are measured with Electroencephalography (EEG). The EEG machine will record brain wave activity by the placement of electrode sensors around the human scalp to pick up brain activity.
It's already established that the Big Brain has similar brain waves depending on the Brain's block function and clock frequency. It would be nice to also have a proper EEG machine to measure or simply observe qualitatively the changes in Big Brain wave activity. Now we have such a device.
The Big Brain has its own EEG machine. You can construct this cheap EEG machine using a very simple technique.
Take a portable Amplitude Modulation radio and tune it the low end of the band at a location without any radio stations. Now place it near the Big Brain. Retune the frequency to a gain of strongest signal, i.e. a brain wave signal sound coming from the Big Brain.
As the DIY EEG is moved around the Brain, different sound will result from different clock frequency. Now you can hear and study brain waves emanating from the Big Brain. Clever interpretation of this sound will lead one to recognize what activity is occurring in the Big Brain.
Brain activity can vary depending on locations inside the Brain. Moving the EEG to different parts of the Brain will pick up this brain wave activity. A more sophisticated model of the EEG radio could quantitatively record Big Brain Waves and their locations. Radio, RFI probes placed around the Big Brain can pick up this activity and produce a 3D model of the Big Brain's activity.
http://www.doctorhugo.org/brainwaves/brainwaves.html
http://en.wikipedia.org/wiki/Electroencephalography
http://blog.makezine.com/archive/2008/11/the-brain-machine.html
http://www.google.com/search?hl=en&newwindow=1&q=brain+waves&aq=f&aqi=&aql=&oq=
Edit: it's also possible to create an EEG from a single Propeller chip. Consult the threads about prop radio.
In a Big Brain circuit with hundreds of Propeller processors, it may be challenging to quickly diagnose if one chip loses its functioning. Now we've never had any Propeller chip fail and we don't expect any to fail in the future, but there is always the possibility of a wire coming loose and shutting down a prop's functioning. Finding that one prop could be like looking for a needle in a haystack.
By using the Qualitative Big Brain EEG Machine (placed at a more remote distance) and a single probe (on the end of a shielded cable) to quickly scan each Propeller chip, a determination is quickly made as to the operating status of the chip.
A program in SPIN to generate some calculations or a simple loop (REPEAT) would be effective wave generator code for testing Propeller chips inside the Big Brain.
Perhaps a neural injection program could also be adapted to specific Propellers in blocks by injecting neurons instructed to fire at a specific speed.
Background
An artificial neuron is a device with many inputs and one output. The neuron has two modes of operation; the training mode and the using mode. In the training mode, the neuron can be trained to fire (or not), for particular input patterns. In the using mode, when a taught input pattern is detected at the input, its associated output becomes the current output. If the input pattern does not belong in the taught list of input patterns, the firing rule is used to determine whether to fire or not. http://www.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/cs11/report.html
Types of neurons - By connection: There are three classes of neurons: afferent neurons, efferent neurons, and interneurons. Afferent neurons carry information from tissues and organs into the central nervous system. Efferent neurons transport signals from the central nervous system to the effector cells. Interneurons connect neurons within the central nervous system. By function: Sensory neurons carry signals from sense organs to the spinal cord and brain. Relay neurons carry messages from one part of the CNS. Motor neurons carry signals from the CNS to muscles, motor neurons are connected to the relay neurons. The signal passes between the neurons via synapses. Synapses are microscopic voids between cells where chemicals are released from the axon terminal of one cell to specialized chemical receptors on the dendrite of the receiving cell. http://simple.wikipedia.org/wiki/Neuron
At this early machine Brain stage, the fabricated machine test neuron is in a Using Mode with a specific firing pattern that can be varied by programming. The "taught input pattern" is usually a test constant of periodicity, simply made into output on pins. To indicate the test firing and the functioning of the firing in each Propeller chip, fire is indicated by output states. These outputs are seen in some of the early photos in on or off states. The example of the working machine neuron is shown in the illustration below. For the Big Brain, the INIT Neuron (a type of machine neuron for testing) is created. This holds the inputs constant as seen to the left in the diagram and gives several things. One, it produces a single neural matter electronic cell that can be propagated to the array. Two, it gives a periodicity of firing feedback from the Propellers. Three, it indicates the INIT Neuron is firing or not firing. INIT Neurons have four states: constant Use Mode, Fire, Negate Fire, Periodicity.
Simple machine Neuron example
http://www.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/cs11/report.html
-Phil
http://www.scilab.org/contrib/index_contrib.php?page=displayContribution&fileID=166
It's rather old code. I've just downloaded it and I'll try it out.
Thanks! When I get back today, I'll look at it in more detail.
There's a link with 130 managed projects that look very interesting found here at the root:
http://forge.scilab.org/
Partial Sample list below