Knowing the structure of your model and providing it training data is a far cry from understanding how it reaches it's conclusion.

> (there can be millions or more)

You just described how incredibly complicated a system can be, yet still attempt to argue my point about programmers not fully understanding the logic behind them.

​

> for each simulated neuron

It's fascinating that you would analogize the way it functions as imitating a neuron, then only later state that everything it can do, could be achieved by a calculator.

​

I don't think you and I will ever agree on this topic. You seem impossibly married to the idea that every single computer is analogous to a calculator. I view that argument as both ignorant and reductive. All attempts I've made haven't produced new arguments from you, but are instead met with heels in sand.

Still appreciate you taking the time to respond, I just don't see this as being a productive use of either of our time.