[alife] Product Announcement - Brain Circuit Simulation Software Demo

[David Olmsted]

I am pleased to announce the free public release of the demo/viewer version of my Brain Circuit Simulator software
for the PC along with many sample circuits. The user is able to create and simulate realistic, adaptive, action 
potential using, neuronal circuits in a 3-D space.

YouTube videos of circuits being simulated can be found at: http://www.youtube.com/user/Ironstead.

The site for downloading the software and the sample circuits is at http://www.softstatemagic.com.

Some sample circuits available for download that can be tested and tweaked include:

    1. Classical Conditioning
    2. Two forms of Pattern Generator including one that transitions to pulses from a constant action potential frequency. 
       These are also a good simple example of neuromodulation in which fast acting adaptive parameters are changed to 
       change the characteristics of the neurons (synaptic weights are not the only ones!)
    3. A Pattern Regularity Detector (Novelty Detector) able to learn patterns without supervision, without separate 
       training sessions, and at any learning speed.
    4. An Adaptive Switching Circuit that accomplishes Operant Conditioning (reinforcement learning with an error 
       criterion or motivation value). Using this circuit as the second stage to a first stage Pattern Regularity 
       Detector allows the arbitrary classification of patterns.
    5. A demonstration of cross-modal enhancement (a dynamic multiplication effect),
    6. A demonstration of time delay
    7. A Decision Gate (winner-take-all network)

This demo software allows you to do everything the full version can do except save to disk. Thus you can create you own 
brain circuits or modify those you download. It also allows you to create time varying analog signals for circuit testing. 
Future versions will take advantage of multiple processor cores (and then computer clusters) and have a 3-D virtual world 
where a simple animal can interact with its environment allowing complete, if simple, brain testing at the system level as 
researchers follow the course of evolution with ever more complex brain systems.

The equations behind the simulations come from the new Soft State Automata theory which describes how non-clocked discontinuous 
systems (like the brain) work. Finite State Automata (binary logic) which is the basis of digital (thus clocked) machines is a 
subset of this theory as are Multivalued (Fuzzy) logic networks and Probabilistic based networks. The brain is biology's attempt 
at implementing a Soft State Automaton. The paper describing this theory is at softstatemagic.com. Needless to say these 
equations can be simulated much more efficiently than can equations based directly on the biology. In addition they also show 
what neural parameters can be controlled.

The web site at softstatemagic.com is not only meant to be a support site for this and future versions of this software but 
also a place where researchers and enthusiasts from all over the world can share and comment on the circuits they develop. 
Advances in this new field will come faster if researchers are able to build upon each others work instead of having to "start 
for scratch" each time.

For those of you in Academia there is a lot of publication potential just from the sample circuits as their limits are tested 
and improvements suggested. Science often advances in proportion to the sophistication of the available tools and this software 
tool has a lot of room for improvement so I need to work on that. The full version will hopefully be available for purchase soon 
at a cost of $39.95.

Sincerely,
David Olmsted