From: Hal Finney (hal@finney.org)
Date: Thu Mar 28 2002 - 11:29:30 MST
Scerir writes:
> 'In principle' an infinite amount of information can be coded
> in a spin 1/2 system (that is - but I'm not sure - proved
> by quantum computation). That's because each point on the
> 'Poincare sphere' (and they are infinite) is a quantum state,
> for a spin 1/2 system. Thus a spin 1/2 quantum system has
> an infinite number of 'possible' (which does not mean
> 'accessible') states.
Again I object specifically to the claim that this information 'can be
coded' into the spin 1/2 system. I take this statement to be a claim
that a certain physical process is possible: that you can start with
a system holding more than 1 bit, let it interact reversibly with the
spin 1/2 system, and at the end of that time the initial system has
been reset to a standard initial stated (erased) and the information
transferred into the spin 1/2 system.
To me, that is what it means to 'code' information into a system, in the
quantum sense. You are transferring the information into the spin 1/2
system. If the spin 1/2 system can really 'code' that much information,
it ought to be possible to transfer that much information into it.
I think you agree that this process is not possible. If it were possible,
then you could get more than one bit out of the spin 1/2 system, which
we know cannot happen.
As for quantum computation, it has many amazing properties, but you
will never be able to physically store more than N bits of information
in N qubits.
Hal
This archive was generated by hypermail 2.1.5 : Sat Nov 02 2002 - 09:13:07 MST