Tom,
"The authors assume that nonlinear positive feedback (noise) in a quantum system can be mitigated by linear negative feedback on the assumption that quantum theory is "fundamentally random." Because of this assumption, the experimental controls on Bell-Aspect type results are statistically reconciled with the theoretical result (Bell-CHSH inequalities) by the imposition of negative feedback from the *experimenter.* In other words, by disallowing nonlinear positive feedback in the experimental protocol, the experiments validating Bell's theorem beg their own conclusion.
Joy's result shows unambiguously that natural nonlinear random input results in the smooth function of quantum correlations, with no boundary between quantum and classical domains."
And from your message to Akinbo:
"...a digital computer cannot do other than compute digitally. Every computer simulation of a continuous function substitutes difference equations for the differential equations that describe a continuum.
An analog computer can faithfully replicate a continuous trajectory, given boundary conditions -- that is not adequate, however, to simulate a continuous measurement function with randomly changing boundary conditions.
Therefore, a continuous measurement function simulated on a digital computer requires random input to a continuous trajectory, in order to replicate the function."
I haven't designed a computer, but, given equal efforts at best quality of construction, I would trust the digital computer over the analog computer to more faithfully replicate a continuous function. My reason for saying this is that I envision that the digital computer code does not deteriorate unless used. And, it is used intermittently. The uncoded analog signal experiences continuous deterioration.
The important question of this messages has to do with:
"... to simulate a continuous measurement function with randomly changing boundary conditions.
Therefore, a continuous measurement function ... requires random input to a continuous trajectory, in order to replicate the function."
I removed "simulated on a digital computer", was it required in some way for your statement to reflect accurately your meaning?
Also, my impression of "...a continuous measurement function with randomly changing boundary conditions...(simulation)...requires random input to a continuous trajectory, in order to replicate the function." is that it is not easily clear to me that this should be true.
A little more explanation might help.
James Putnam