Download, double click, move mouse to randomize and vibrate any part of the pixel/point/neuronlike/partialqubit grid, and watch heat death (if you randomize enough to kill the discrete patterns that emerge when time is set to a positive number in the code as by default) get darker until, and this is the same equation operating the whole time not seperate modes, a mix of discrete and wave-like patterns emerge and in some ways look like brain waves, since I did use some basics of Kurzweil's new AI book to figure out how to implement time as the inverse function of gravity, just add and subract a tiny number from each very very accurately percentile fraction balanced node and edge and nodes and edges near them. This is the real deal. It learns to form and move basic shapes but I didn't tell it how. It starts forming more complex patterns completely on its own. There's no code or equations in there saying to do that, except if you can imagine the math of how increasing weights between nodes as a reaction to 2 nodes having opposite numbers, or decreasing for equal numbers (and everywhere between gradually), that it forms a structure of points in space similar to layers of neocortex as in Ray Kurzweil's (who recently became Google's director of engineering) and Jeff Hawkins' theories. I've now expanded those theories to show a similarity between brains and physics with a fun simulation you can play with using the mouse. Try it and write what you think.
https://sourceforge.net/projects/physicsmataWhat should also interest members of this forum is that when this changes the world, if people can understand what its for and start building things, both physics and artificial intelligence at least, with the public domain 24 kilobyte (much larger than it should be but I was in a rush to get it working at all) executable jar file (and unzip to get source code), then there are big things ahead for everyone, depending on what futures we as a species prefer and can agree on plans toward. Warp drive is no longer to be laughed at. We have accuracy to seriously talk about low power warp drive mass produced, eventually when we learn to do things safely and not crash our warp drives into buildings or planets while dragging asteroids behind us, for example. More short term, and more directly related to this forum, might our thousands of dollars worth of Bitcoins start to be worth millions or billions a short time after people realize the potential futures we have available in, not just what I did but FoldIt, RepRap, Wikipedia, and many other things we can build into our preferred futures when we decide to work directly together like I'm doing instead of letting our leaders have all the fun. How much do you think Bitcoin's value would increase in such events? And how can we get that to happen sooner or what other ways could it be done?
If you read about my early PredictionVoxel plans, which were part of what led to this, I am interested in minimalist small granularity of prediction to organize the world better where money is more like variables and math than something that can be regulated. I'm into new kinds of economics. Anyone interested? So many tools available these days, its hard to decide what to build next.
Here's what I wrote about it at kurzweil's forum
http://www.kurzweilai.net/forums/topic/try-this-singularity-prototype-only-144-neurons-and-already-moving-simple-shapesThe algorithm had to be simple, equation simple, and it is. Recently I've been writing some crazy sounding things, but its not easy trying to understand the nature of the universe and how math and science and many things can fit together into a global network which I hope we will build together from this public domain (anything goes) file you can double click to run right away or unzip to get source code, a file which is now intelligent and you can paint onto with the mouse in a basic way to see what Google does not know how to do (they are welcome to the code like everyone else), what has never been done before, just enough to show you what its for and how to build a real global system. What we have now is just barely fine-tuned and varies in behavior by timing as thats how I defined how much forward to calculate and other subtleties, but it works in any case. It works better as I've given you the file. Don't overlook any detail which may contribute to its subtle behaviors, like that I did not update all the nodes and edges at once but let it overlap without buffering and it appears to help like a tiny twist of quantum wave packet compared to a balance in that part. Most importantly, it does nothing useful or interesting without the line to make it .000001 tiny fraction out of balance from the very well designed normalizing I did as percentile fractions applied at each node and edge and its surrounding nodes and edges. Without that tiny imbalance, its a timeless simulation and sits very still but you can still draw on it. Turn it negative and time is replaced by gravity. Its a quantum gravity simulation, specificly the very abstract part which telepathy and the Law Of Attraction (attraction is gravity? who would have thought?) use in recursive patterns and in networks of patterns with cycles, to mentally find paths through the space of multiverse possibilities and back to a slightly different state of the world or telepathic connection, for example. But more importantly, just remember that the recursive self referencing self improving intelligence, even with just 144 neurons so far, works because its a simulation of quantum gravity and the thoughts gravitate toward eachother in the context of kolmogorov complexity and degrees of freedom in the space of possible states of the pixels on screen which are each a neuron that varies from 0.0 to 1.0 and is normalized by percentile fraction as are the edges too.
Just go play with it. You'll see what I mean. Then spread the word that we have the code we need to build what we've only talked about for too long and as a small and simple code.
Version 1.0, public domain (you can't patent the laws of physics), download then double click.
https://sourceforge.net/projects/physicsmata 
