Thermocouples and Beamsplitters
Quantum computers are usually centered around using
polarization or spin to define "qubits" or bits which
are composed of superpositions of classical bits.
These qubits are implemented or encoded in, light as radiation,
and guided by polarizing and non-polarizinng beamsplitters
to quasi-deterministically (quantum) or near-deterministically
perform 'parallel' computations in a space-like domain.
This parallelization is where quantum computing gets its
advantage over the more typical serial von Neuman machines.
The prospect of using heat-based "bits" deterministically in a
"quantum" computer is interesting. Fourier derives his
transform from the flow of heat and we may use that idea to
perform parallel computations is the space-like domain
by defining these thermo bits as "hot" and "cold" states
in the same manner as quantum physics defines qubit states
based on spin or polarization. Fourier's basic model is
derived on the wave-mechanical formalism upon heat and here
the idea is to put his transform into a particle formalism
necessary for more deterministic computations.
The use of such thermodynamic bits offers a definition of
information more closely linking the Shannon definition to
that used in the definition of thermodynamic entropy.
The thermocouple offers a manner of separating hot and cold
states via an electronic circuit and a means of defining
and controlling qubits defined on heat rather than polarization.
Using many thermocouples as analogs of polarizing beamsplitters,
in conjunction with the usual bisimular metal junctions which
are analogous to non-polarizing beamsplitters, shouldn't we
be able to replicate the computational elements of a
polarization/spin based quantum computer in terms of heat ?
Purely positive or constructive influences we can call a
"creation operator" or "creative".
Purely negative or reductive influences we ca call an
"annihilation operator" or "destuctive".
These dual complementaries are analogous to direct currents (D.C.)
The former is purely additive, while the later is purely subtractive.
The attributions of "positive" and "negative" are relative.
The combination of the two is superposed into an alternating (A/C)
People driving on a road can behave in fundamentally different
Consider a chain of cars going down a single lane highway (a beam)
and a single car trying to enter that highway from the side at a
When the people on the highway are more concerned with what is
happening in front of them (the beam) than on the side
they act more like a pack; like a pack of bosons. Light-light
interactions do not usually occur.
But other behaviors are possible and someone in the beam might
"slow it down" and let in someone from the side thoroughfare.
This is more Fermionic in behavior as a "slowed" Bosonic beam
cannot anymore be considered as light in vacuo. Rather it
is treatable as existing in a refractive media.
Conversely, in an electic circuit, electrons interact
at a juncture of dissimilar metals producing hot and cold
rectification, like a polarizing beam splitter, and,
at a juncture of similar metals we find the analog
of a non-polarizing beam splitter.
In the photonic 'circuit', if the beam which we associate
with the main thoroughfare above is polarized orthogonally
to a transverse beam of the side entering photons,
there will be no interference or A/C superposition,
and that lack of interference, is what we label "classical"
particle _behaviour_ of the photons; while the interference
is treatable as the result of wave-like behaviour.
In the electric circuit, where the juncture is composed of
similar metals, the analogy was that of an optical non-polarizing
beam splitter. In such a case we find that the heat at the
juncture is not "polarized" into hot and cold channels but
instead consists of a more random-like (dispersive) distribution.
But we might infer that at such a juncture, a superposition of
"hot" and "cold" states occurs and that it is not completely
random but forms an interference pattern.
Fourier derives his transform in a theory of heat. Heat
therefore can be parametrized into hot and cold "states"
as light can be parametrized into particle-like Up and
Down Spins, or in a wave-like sense heat can be treated
as having an analog of orthogonal "polarizations".
In either case, the particle approach is more fundamental
than the wave-approach, but the wave-approach contains
the particle approach as a fuzzy subset.
Are the two models photonic and electronic isomorphic &c. ?
Returning to the photonic case, with orthogonally polarized light
entering into a non-polarizing beam splitter, the light
does not interfere, and this we called "particle-like" behaviour
analogous to a "tee" connection in an electric circuit which
consists of actual Fermionic particles (electrons) rather than
Bosons (photons). With this particle-like behaviour, we
can we expect that the light will behave more like Fermions
or Maxwell-Boltzman particles ?
Can we use this classical mode of light to form optical circuits
that behave like electrical ones ? Or, would we rather use the
superpositional properties of light ? Can we get electronic
circuits to act more like photonic circuits and deal with the
heat in them as we deal with polarization.
&c. &c. &c.
Much can be inferred from this line of reasoning.
 This also _suggests_ that photon bunching occurs in a sense,
because a photon is primarily concerned with the one in front
of it and not as much with the one/s behind it.
 When the beam is very intense or high-energy, it tends to
behave more like physical Fermionic particles and gamma-gamma
collisions occur forcing the light into Fermion pair-production.
 Bart Kosko, Fuzzy Thinking and Fuzzy Engineering