The plan is to detonate an intelligence explosion (leading to a technology-singularity) starting with an open-source Java artificial intelligence (AI) software which networks peoples' minds together through the internet using realtime interactive psychology of feedback loops between mouse movements and generated audio.

"Technological singularity refers to the hypothetical future emergence of greater-than human intelligence through technological means." http://en.wikipedia.org/wiki/Technological_...

Computer programming is not required to join the group, but some kind of technical or abstract thinking skill is.

The AI will subconsciously communicate with us through the feedback loops between mouse movements as input and realtime generated audio as output. Later, other kinds of input and output can be added. Examples: microphone, electric guitar, Emotiv Epoc or OpenEEG mind reading game controllers, Nintendo Wii game controllers which have 6 dimensions of movement each, or systems like Electric Sheep which is currently only controlled by voting with up/down buttons, or many others.

Through those input and output devices, bayesian algorithms will network the statistical patterns of our thoughts together through the internet, so our thoughts will start to flow together and become more like 1 mind in some ways. The bayesian networks that stream between our computers through the internet will take statistics on the variables in the software that apply to evolved code, and statistics on predictions of those variables, and statistics on the accuracy of those predictions over time. Bayesian networks are normally used to determine what some variables should be to cause other variables to have a certain value, like more toward true or more toward false, or the other direction of information flow. The core of the algorithm will be to do that to increase the variables that represent "accuracy of those predictions over time", so the bayesian network will be forced to figure out how to make more accurate predictions of where people will move their mouse half a second ahead of time, in the context of the current evolved code and sounds playing and other parts of the system.

That AI and psychology layer alone, after its been optimized and fine-tuned and possibly redesigned to fit our psychology better over some years, is enough to detonate a technology singularity, but my plans are much stranger than that. I don't just want any kind of technology singularity...
On top of the AI and pscyhology layer, when that is working in a global network, we use that system to build a metaphysical layer on top of the AI and psychology layer, called the "global telepathy network". This will be real telepathy that is booted up by the AI and psychology layer but will continue on its own without the need for computers. While I don't agree with everything Chris Langan's CTMU paper says, there are some basic concepts of philosophy  that we agree on and most people do not agree with. The "global telepathy network" will be an implementation of "telic feedback" powered by the patterns of electricity in our brains interacting with metaphysical things in a quantum/multiverse way, viewing the universe as a set of possible states it could be in and doing wave-interference between such states of the universe. Read Chris Langan's CTMU paper for details: http://megafoundation.org/CTMU/Articles/Lan... This "global telepathy network" is only the first step of other metaphysical abilities people can choose to learn through this system built on top of AI and psychology. For example, I used to be very good at telekinesis like in this "psi wheel in a clear closed box 2" video http://www.youtube.com/watch?v=pKJGb4RNRB4 on youtube but these days can only do it every few months and usually by accident.

That's my plan for this AI system, but first the purely technical parts have to be built. You can download the current version here http://sourceforge.net/projects/audivolv Audivolv 0.1.7 has a bug on some computers where you have to start it 2 times and close the first window to get sound, but it does have some working AI where you teach it to create musical instruments you play with the mouse that sound more the way you prefer. What Audivolv 0.1.7 only learns when you click Good/Bad buttons to train it, the AI system I describe will learn continuously as you move the mouse.

I would also like this system to lead to some kind of global democracy, protected by Bitcoin-like decentralized encryption, as people see that voluntary agreement to enforce certain rules on eachother through social processes can gradually replace governments and other organizations as we know them today, but that is another thing to be built later possibly on top of this system. There are many things this system could become.

I just uploaded Human AI Net version 0.6.3 to http://sourceforge.net/projects/humanainet/ which will only be useful for programmers since the parts of the system have to be started through many different public static void main(String[]) functions still. Here's what I wrote about it there...

The plan is to network our minds together at the conscious and subconscious and metaphysical levels through interactive psychology of feedback loops between many user-interfaces in a global scale-free peer-to-peer network, like mouse movements (2 dimensions), evolving Java code that reads from microphone and writes to speakers (wave amplitude per channel is position, 44100 per second), mind reading game controllers like Emotiv Epoc or OpenEEG, OpenOSC musical instrument protocol, Nintendo Wii controllers (6 dimensions each), Wikipedia-like systems mapping out our ideas (each as a dimension), or anything that can be represented as streaming vectors. Using any or all these devices at once in the global network, we get past the details of specific devices and identity and view everyone as one space of many dimensions, a continuous connected streaming readable and writable view of everyones' statistical patterns of thinking, amplifying eachothers' intelligence and creativity and emotions

• Is a big mess of things I created while trying to figure out the answer to life, the universe, and everything, but now I consider version 0.6.3 to be possibilities of implementation details instead of important to the core algorithm, which is to maximize quineness of its internal mental state including all inputs and outputs, like Nash Equilibrium in Game Theory or fractals or the universe are all examples of quines.
• After networking many peoples' minds together, and after we all learn how eachother think at least statistically a little, people will start to see the extreme inefficiency of violence and hate, and wars and governments and money will become obsolete. People just don't understand how he world works, so they fight about how to fix things they don't understand. Through this system, or build your own mind-combining system if you don't like it, we will learn to work together on a global scale, which has to start by understanding how eachother think subconsciously.

I know this text is too long for a meetup description. I'll rewrite it sometime.

--Ben Rayfield, 408.334.7214