When I was invited to do this, I was told you were interested in:
"alternative methods, which, eventually, could result in some kind of collaboration. They are indeed interestd in learning more about... various NN models."
So, I'm going to tell you a bit about real neural models (which I don't think your particularly interested in), and a bit more about Deep Nets.
I'm going to say a bit about No Free Lunch Theorems, and why they're good for deep nets.
Finally, I'm going to tell you a bit about Middlesex AI hoping to get some collaboration going.