NeurOSelf

Integrating OpenCog and Critterding

2010-01-16T15:16:00.000-08:00

Once Critterding critter controllers are abstract and modular then we can drop-in an OpenCog instance - as an alternative to the default critter BRAINZ neural network... or some hybrid combination.

NeurOSelf First look

2010-01-07T23:57:00.000-08:00

Twitter Operating System: NeurOSelf's Twitter Interface

2010-01-05T10:29:00.000-08:00

The success of the internet can be traced to the interactivity it has given people to each other, already some of the most successful websites on the internet are those that bring together the most people. Wikipedia for example is currently ranked as one of the most successful websites on the internet, and it's functionality is basically creating a platform for people to collaborate the thought on an array of topics. This interconnectivity between users will not just go away in my opinion but will actually increase as more technologies are created to interact with the collective intelligence of the planet.

One particular tool that is currently growing is Twitter which has the potential to become more than just a communication tool. Twitter is still in an early infant state, and because of this we believe that Twitter has the potential to grow into a significant hivemind component of Earth's internet brain for the forseeable future. At this point, Twitter's potentials are huge and relatively unexplored.

The interface shared by both the Twitter.com and advanced "desktop" clients such as TweetDeck are begging for a drastic improvement: a system of adaptive filters that manage the presentation of a potentially large, loud, and noisy "firehoses" of streaming data being pumped and replicated across the system. These will involve more than simple on/off switches: they will resemble interactive "graphical equalizers" of data that can smoothly tune in-to and out-of certain "frequencies" or features of incoming, ambient, and archived media streams. Already, services like Tweetdeck and Twittree attempt to provide new ways to organize data. Yet We believe all of the present systems are relatively archaic in comparison to how the internet is evolving. We propose its time for a revolution, a revolution in human-computer interfaces and data experiences.

Why a neural network interface?

Imagine being in a room with 30 people and everyone is having a random walk conversation with everybody else. It becomes nearly impossible to follow a thoughts development through the room if you can't map out the interactivity the people have with each other, this is why I believe the best solution to this is mapping out all the interactions in a fractal space - 2D or 3D, or a combination of both. This allows objects to be scaled arbitrarily, relative to eachother. A single thought or a twit ("thit"=spontaneous thoughts expressed as twits) can be presented as part of a neural web that interlinks them in all meaningful ways. This level of interactivity not only gives you the ability to see how people are interacting with any particular thought but how others are interacting with their thoughts about your thoughts.

Central Brain Units (CBU) - Global Brain Paths to the Singularity

2009-12-29T18:31:00.000-08:00

Welcome to Web 3.0 - the web that not only connects us, but an internet that also acts as a massive collaborative tool. We will begin our exploration in the creation of the first cloud - what I call a CBU (Central Brain Unit). CBUs are essentially computer social networks like MySpace, Facebook, YouTube, Twitter, or something else that has yet to emerge. Unlike current social networks, CBUs contain tools that enable a massive amount of people to collectively and simultaneously work on problems, play games, create organizations, or achieve any task that serves a conceivable function or purpose.

Imagine a tool that gives hundreds of people the ability to collaboratively draw a single picture. At first you would think that it would be too chaotic with people purposely not wanting to participate constructively. Yet I believe those peoples' actions would ultimately be self regulated by the collective network. Everyone will be responsible for not just drawing, but also deleting segments to the drawing that they believe are not constructive to the overall picture. The CBU's would cover huge amounts of specialized networks such as art, education, entertainment, or data collection and organization.

Yes, a CBU would act much like a wiki or wave but with one major difference: it would leverage the power of crowds that social networks tap into, and focus it towards specific tasks. Imagine a wikipedia entry with several million people adding information to it simultaneously -- this massive collaborative network would be at the heart of what I believe web 3.0 will be all about.

CBU of 2 to 1,000,000 users

Over time, each of these specialized networks would collaborate information with other CBU's and other networks with massive amounts of users involved in specialized tasks. This greater cluster of focused intelligence I will call a SBU (Super Brain Unit).

SBU of 1 million to 100 million users

Repeat the same process as above, but now the SBU's are connected to form what I call a GBU (Grand Brain Unit)

GBU of 100 million - 1 billion users

Once a GBU forms, we may be able to focus intelligence enough so that synergistic super-intelligence emerges - causing a runaway singularity effect.

Technicalities

The creation of the first CBU may begin through current web based internet social technologies. Over time, as GBU's form, the actual interconnectivity between the brains may automatically become more dense, internally. There is no technical need to have the brains connected with conventional web interfaces: it may actually be greatly more efficient to connect the brains through some form of neural net.

How and when the connectivity changes may have a direct effect in the acceleration rate towards an apparent singularity situation.

Emergence of Super Intelligence

There are several reasons why I believe the creation of a GBU will lead to the emergence of super intelligence. I will briefly and simply summarize some of the possibilities I am imagining:

The GBU acts as a collective consciousness, within itself.
The GBU may learn how to make CBUs more powerful creating an exponentially rise in collective intelligence.
Thus the GBU may become infinitely recursive Ai software as the ultimate result of our interactions inside it
The GBU, as a software scientist, may solve theorems and discover patterns leading to advances in science (such as physics and biology) - that are not yet known or even understandable by humans.

The Uploading Renaissance

2009-12-29T18:27:00.000-08:00

The Uploading Renaissance (as I have dubbed it) marks a very pivotal period in the way we interact with knowledge, data, and information. Currently the brain is disconnected from data by its senses: If you wish to know the information contained inside a book, you must go through your eye input to recreate the information. Of course you can also load the same information via audio or touch. With technologies that convert text-to-speech, blind people also can access the information contained in books. But what if you are blind and deaf? Well, we also have a technology called braille in which one feels information patterns with finger tips. All of this is possible by the extremely diverse modalities supported by the nervous system that convert sensory signals into information, data, facts, knowledge, and ultimately wisdom. Yet some currently argue that all these methods are actually slowing us down. The days of reading information is a thing of the past with the creation of ever-improving knowledge uploading technologies. Imagine downloading a recipe for a cake that you tasted in a neural simulation - this and other such possibility open up a world for an explosion in the overall potential knowledge capacity of the human experience. Complex math and abstract philosophy may one day be available for upload directly into the brain, bypassing the lag inherent in your senses, at least this is what is believed would happen in theory.

http://www.youtube.com/watch?v=l7_IklNCYAI

The strange truth is that theory seems to be outpacing actual biological knowledge of the feasiblity of such technologies. Rendering all of these concepts short of experiments of the mad scientist.

Open-Source Neural Simulation Software

2009-12-29T17:41:00.000-08:00

This article deals with a possible path to the creation of a virtual global brain.

We currently live in a world where any computer software functionality is copied and replicated within weeks or even days of its release by the open-source market-place of information technologies. But there is a computer program which does not yet have an open source equivalent: the brain simulation software used by IBM in its Blue Brain research project. This, like many other software projects of its kind, is surely to be created within a short period of time and it's potential for launching a singularity may surpass that of its current, isolated "research lab" incarnation.

Neural simulation and other connectionist technologies form the expanding frontier of the field of Artificial Intelligence - both emulating the brain, and distilling its elegant essence for maximum computational efficiency. The potential expansions from these technologies, while not presently known, may have a similar effect as the Human Genome Project's explosion of consumer-level genetic technology.

Yet the creation of an open source equivalent to such a program to this program and other similar to it that will soon be created will some of the most potentially disruptive applications to hit the internet since Google. I believe that an open source equivalent to open to the blue brain can be created now with way much less line of code that most people believe it will take. We don’t need to code an entire brain to create a massively parallel system, I believe this can be achieved by a program similar to SETI@Home which brings together millions of computers together to serve a single function.

I know you may be thinking that networked computers may not catch up to the supercomputer efforts, yet I want to remind you that the first computer to reach a petaflop was a collection of thousands of playstation 3 system running the Stanford folding@home program. This extreme flexibility I believe will have the potential to increase the speed of brain emulation software developement and progress, thus creating a feed back loop in the artificial intelligence and brain simulation industry,rendering the open source

While it may not be clear when such a program may appear on the internet, I do believe it is only a matter of time before such a thing is created, thus driving the the needed debated associated of wheter we want the internet to essentially turn into artificial intelligence.

While it is uncertain how long research such as this will take to fully map out the human brain we can expect this to only increase in exponential growth alongside the overall growth of information technologies. Therefore it may lead to the actual simulation of a human brain in silicon within a decade or two, assuming we don’t experience an extreme paradigm shift in computer architecture or an actual shift in our approach to computing.

(Realistic) Global Goal: One neuron per CPU core, forming a massively parallel web 3.0 central brain unit.

Caution:

It may not be possible to retain control on a system that has been created and distrubuted by many users on the internet thus rendering a physical cut of switch to the internet brain a near impossible task.
Personal computer no longer "personal" - to the degree of boundary dissolution decided by the computer operator.
Once a neural brain world wide operating system is engaged, it may not be possible to keep any piece of information private

Upsides:

Unexpected surprises that may emerge from the patient chaos of a well-connected, increasingly dense virtual nervous system.
Owned and guided by the people who operate the computers of which it is comprised - not a private or centralized facility.

List of Neural Simulators (as of 29 Dec 2009):

from: http://grey.colorado.edu/emergent/index.php/Comparison_of_Neural_Network_Simulators

Simulator	Publications	Version (as of this writing)	First release	Latest release	Primary author	License	MS Windows	Mac OS X	Linux	Other	Active Community	Language
Emergent (formerlyPDP++ and PDP)	AisaMingusOReilly07	5.0.1	1986	2009	Dr. Randy O'Reilly	GNU GPL	XP, 2003, Vista	Intel, PPC	Any, Fedora, Ubuntu	Any Unix	emergent-users list, Wiki	C++
FANN (Fast Artificial Neural Network Library)	Nissen03	2.1.0b	2004	2006	Steffen Nissen	GNU LGPL	Yes	Yes	Yes	Any Unix	FANN Forum	C
GENESIS (the GEneral NEural SImulation System)	BeemanEtAl07	2.3	1988	2007	Dr. James Bower &Dr. Dave Beeman	GNU GPL	Cygwin	Intel, PPC	Yes	Any Unix	SourceForge list	C
NEURON (originally CABLE)	Hines93 HinesCarnevale97 HinesEtAl06 See full list athttp://www.neuron.yale.edu/neuron/nrnpubs	7.1	1986	2009	Dr. Michael Hines	GNU GPL	95, 98, NT, 2000, XP, Vista	Intel, PPC	Any Linux	Any Unix	NEURON Forum	C, C++, Fortran
Neuron-C	Smith92	6.44	1986	2008	Dr. Robert Smith	GNU GPL				Any Unix		C,C++
SNNAP (Simulator for Neural Networks and Action Potentials)	Unknown	8.1	2001	2007	Dr. John Byrne & Dr. Douglas Baxter	©, Proprietary	Java	Java	Java	Java	Available but defunct	Java
LENS (The light, efficient neural network simulator)	Rohde99	2.63	1998	2000	Dr. Douglas Rohde	Unknown	Cygwin	Intel, PPC	Yes	Any Unix	No	C
Catacomb2 (Components And Tools for Accessible COmputer Modeling in Biology	Unknown	2.111	2001	2003	Robert Cannon	GNU GPL	Java	Java	Java	Java	No	Java
Topographica Neural Map Simulator	BednarEtAl04	0.9.4	1998	2008	Dr. James A. Bednar	GNU GPL	Vista, XP, NT	Build from source	Build from source	Build from source	Mailing list, boards	Python/C++
(iNVT) iLab Neuromorphic Vision C++ Toolkit	Unknown	Unknown	2001	2005	Dr. Laurent Itti & Dr. Christof Koch	GNU GPL Must request access	Unknown	Unknown	Unknown	Unknown	iLab forums	C++
NEST (NEural Simulation Tool)	DiesmannEtAl95 DiesmannGewaltig02 GewaltigEtAl02 EpplerEtAl08	2.0 beta	2004	2009	NEST Initiative	©, Proprietary	Build from source	Build from source	Build from source	Sun Solaris HP Tru64 Cygwin SGI Altix	NEST-users list	Python/C++
KInNeSS - KDE Integrated NeuroSimulation Software	GorchotechnikovEtAl04 GrossbergEtAl05	0.3.4	2004	2008	Dr. Anatoli Gorchetchnikov	GNU GPL	No	No	KDE 3.1 required	No	No	C++
XNBC	VibertAzmy92 VibertEtAl97 VibertEtAl01	9.10-h	1988	2006	Dr. Jean-François VIBERT	GNU GPL	9x, 2000, XP	Build from source	RPM (Fedora), Build from source	Tru 64, Ultrix, AIX, SunOS, HPux	No	C++
Stuttgart Neural Network Simulator	ZellEtAl91 Petron99	4.0	1991	2001	Dr. Andreas Zell	Free for academic use Open source	2000	No	Not explicitly	Any Unix, Sparc, DEC, IBM PC, HP 9000, SGI Indigo	Defunct SNNS list	C++
MATLAB Neural Network Toolbox	Unknown	6.0	1998	2008	MathWorks	©, Proprietary	Yes	Yes	Yes	Solaris 64-bit	No
Wolfram Mathematica Neural Networks	Unknown	1.0.2	2002	2003	Wolfram Research Inc.	©, Proprietary	Vista, XP, 2003, 2000, Me	Intel, PPC	Yes	Solaris, Sparc, HP-UX, IBM Power	No
Peltarion Synapse	Unknown		2006	2008	Peltarion	©, Proprietary	2000, XP, Vista	No	No	No	No	.NET
PCSIM: A Parallel neuralCircuit SIMulator	BretteEtAl07	0.5.0	2008	2008	Dr. Pecevski Dejan Dr. Thomas Natschlager	GNU GPL	Under development	Build from source	Build from source	No	Mailing list	Python/C++/Java
Brian	GoodmanBrette2008	1.1	2008	2009	Romain Brette Dan Goodman	GNU GPL	Yes	Yes	Yes	Any OS supporing python and numpy	Support forum	Python

MycoComputer

2009-12-29T11:50:00.000-08:00

Here are a collection of texts, that when combined, describe a mycocomputer - a computer built from mycelium networks and interfaced with electronics.

Microbial "Neural Network": Artificial Intelligence from Fungi (1997)

"Individual mycelia may be considered as neuron-like biological element of NC. The difference between neuron and mycelium is in great number of mycelial exit channels. Natural neural networks and mycelial networks are very similar systems.Taking into account the technological possibility to cultivate the fungal mycelium and to connect them with any electric system,it would be interesting to build AI as a hybrid of FAI and computer-enhanced learning and working system.The mass of one mycelium is about one microgramm and the information capacity of one mycelium as pseudo-neuron is about 10 bits .Thus, the specific volume of FAI memory is about 10,000,000 bits per 1 g of fungal biomass.Ultraweak photon emission of Psilocybe cubensis mycelium tissue and acoustic wave treatments

All kinds of living matter (cell, organ tissue or organism) emit certain amount of light quantum that can be measured as ultraweak photon emission, also called biophoton emission. In this paper we investigate the photon emission of mycelium cultures of Psilocybe cubensis that were treated with acoustic waves of 194.71 Hz for 13 days compared to non-treated control samples."

Mycelium: Earth's Natural Internet by Paul Stamets
"Waves of mycelial networks intersect and permeate one another. This interspersing is the foundation of soils worldwide. Although the mycelia, under the microscope, are seemingly undifferentiated, their ability to respond to natural disasters and sudden changes in the environment is a testament to their inherent intelligence. I believe mycelia are Earth's natural Internet, the essential wiring of the Gaian consciousness. The recent creation of the computer Internet is merely an extension of a successful biological model that has evolved on this planet for billions of years.

The timing of the computer Internet should not be construed as happenstance. Sharing intelligence may be the only way to save endangered ecosystems. The planet is calling out to us. Will we listen in time? The lessons are around us. Will we learn?

Covering most landmasses on the planet, and indeed floating in the oceans, are huge masses of fine filaments of living cells from Fungi, a kingdom barely explored. More than a mile of these cells, called mycelia, can permeate a cubic inch of soil. Fungal mats are now known as the largest biological entities on the planet, with some individual mats covering more than 20,000 acres. The momentum of mycelial mass from a single mushroom species, growing outwards at one-quarter to two inches per day, staggers the imagination. These silent mycelial tsunamis affect all biological systems upon which they are dependent. As one fungus matures and dies back, a panoply of other fungi quickly comes into play. Every ounce of soil hosts not just one species, but literally thousands of species of fungi. Of the estimated 6,000,000 species in the world, we have catalogued only about 50,000. The genetic diversity of fungi is vast by design, and apparently crucial for life to continue."

The Mushroom Speaks by Terence McKenna

"These networks may cover acres and may have far more connections than the number in a human brain.
My mycelial network is nearly immortal--only the sudden toxification of a planet or the explosion of it's parent star can wipe me out. By means impossible to explain because of certain misconceptions in your model of reality all my mycelial networks in the galaxy are in hyper-light communication through space and time.
The mycelial body is as fragile as a spider's web but the collective hypermind and memory is a vast historical archive of the career of evolving intelligence on many worlds in our spiral star swarm. Space, you see, is a vast ocean to those hardy life forms that have the ability to reproduce from spores, for spores are covered with the hardest organic substance known.

Across the aeons of time and space drift many spore forming life-forms in suspended animation for millions of years until contact is made with a suitable environment. Few such species are minded, only myself and my recently evolved near relatives have achieved the hyper-communication mode and memory capacity that makes us leading members in the community of galactic intelligence. How the hyper-communication mode operates is a secret which will not be lightly given to humans."