Re: Quantum Cryptography can not work

Unruh wrote:

Because there *are no hidden variables*. The other photon could *not*
have (always) been in a spin eigenstate before being observed, even
though it *always* turns out to have the opposite spin to that of the
other photon, *if* the same polarization attribute is measured for
both - because of the way the spins turn out when two *different*
polarizations have been measured.

Quantum correlations differ from classical.

Yes, you get to have Bose-Einstein statistics and Fermi-Dirac
statistics in addition to Maxwell-Boltzmann statistics. You can play
at dominoes and cards in addition to dice.

But when I say that the wave function must transfer information
internally, I do mean what I say. This internal information transfer
is not externally exploitable or visible, but its consequences are.

If a standard quantum wave function were *independently* collapsed to
a spin eigenstate at the two locations in a spacelike relation to each
other, then the only way you could have 100% anti-correlation between
the detected spins is if your emitter were producing the particle
pairs in spin eigenstates matching your polarizer to begin with.

If, instead, the quantum state contained "hidden variables", then as
you increased the angle between the two polarizers, the correlation
failure rate would increase linearly with the angle. Instead, it is
much smaller, proportional to the square of the angle.

Thus, one of the two observation effects *must* somehow influence the
outcome of the other; if the two observation effects were wholly
independent, local, and self-contained, nothing could reproduce the
experimental results.

John Savard

.