The point of the post was mostly "people are state machines too" and arguing we don't even know what we are, how will we evaluate AI? Google was giving me actual results every time I searched, and my pattern-matching-happy brain went along with it :D

I spent a few hours trying to build a quantum markov chain. Working theory was that directing the phase when words appear in the same sentence meant interference would cancel out different meanings of a word.

Then I loaded a Dr Seuss book and `cirq` tried to allocate 64 Petabytes. It doesn't even matter at that point.

correlation doesn't imply causation, but it doesn't refute it either.

I guess my underlying point is "what if probabilistic neural networks are an oversimplification"

To use an analogy, what if amplitudes allow for a richer representation of state than probabilities in the same way that neural networks allow for a richer representation of state than markov chains?