I know what stochastic is. I have created stochastic ANNs. I don't think you understand the problem of genetic algorithms with this respect.
There is no problem -- the numbers game is an old and defeated "argument."
I have created stochastic ANNs.
This is at odds with your earlier post. I invite you to repost to consolidate them.
And please post a sample ANN -- I have been a developer for over 30 years and am fluent in about 10 programming languages -- for our purposes meta-code will suffice. Syntax is not an issue -- the algorithmic intent will be clear. If you have the sophistication you intimate I will certainly invite some friends to look at your work.
If your code bears out your contention I will be the first to concede defeat (although what the rest of the millions of physical scientists will do is unknown).