First you need to prove that human cognitive function is also not fundamentally driven by statistical processes.

It's possible that "statistically driven prediction" is all we are.