Yeah, I'm on it more than ever. I've run into a problem, though, which will take some thought: It is possible to create a standard Neural Net as a single class. The class would take the number of inputs, the number of outputs, and the number of hidden layers as arguments to the constructor. The rest can all be done in code. That makes NN kind of easy to get your head around. I have built my Brain Designer, but I have yet to come up with even a vague rule of thumb as to how to use the thing, which means that I can't let anyone use it, yet, because I can't explain what a person is supposed to do with it. Heck, I'm not even sure, yet, myself. So, the result is that I have an idea, and some software based on the idea, and it is pretty cool, but I can't explain it, so it is technically useless.

My next step is to try it out in some scenarios to try to figure out some guidelines as to how to use it. It's an interesting situation. If you have an idea that you can't describe, is it really an idea? Perhaps a concept is only valuable when you are able to communicate it to somebody else.