Computers are real

[Vkothii]

And therefore...?

A computer does "know" things, a computer knows how to run programs, like a lawnmower knows how to mow a lawn.
What about this question, is a pair of sunglasses "a computer"?? Do sunglasses "process" photons, or communicate anything?

What about the entropy in the alphabet of photon polarization?
What's the expectation that each particle, or signal, will be polarized as it impinges on the [outer surface], the input as it were, and either transmitted or reflected/absorbed, by the sunnies?

Start with the incident radiation, a waveform $$ \Psi_{(t)} $$, with all possible polarizations (i.e. along any given x-y plane).

 

[Pete]

But you would fail any question in a physics exam, if you said "information has no mass or energy".

No, I really don't think so. Unless you were the examiner.

Does mass or energy have information instead? Do you still think thoughts aren't real physical things?

No, I never said or thought that thoughts are not real.

P.S. If you ever do study IT, it's one of the first things you learn, even before you get near Shannon's limit. Information has entropy, because it has mass; mass has energy.

Yes, information has entropy. No, not because it mass. Entropy does not imply mass, and vice versa.

I'm not sure you understand what I'm saying Vkothii... but try it this way:

Information is real. Thoughts, ideas, etc are real. They exist. But, they do not have mass or energy.
Information in practice is always associated with things that do have mass/energy, and the information could be considered properties of those things.
But, the mass/energy involved is not a property of the information - it is a property of the medium.

Consider a punch card. One of its properties is its mass, another is its color, another is its position, another is its size, and another is the information stored on it.

Note that all those properties are real things... but none of them themselves have mass as a property.

 

[Pete]

So let me get this straight. If we take a freshly formatted USB-stick and weigh it, then cram it full of data and weigh it again, it has gained in weight according to you ?

Yes, if it has extra electrons as stored charges, then it has gained 'weight', or it now has more particles, organized as discrete bits, in it.
So absolutely, it has more information entropy (which is always independent of the entropy of its physical realization).

That isn't just "according to me", btw.

Take a blank punch card. Weigh it.
Now store a bunch of data on it, and weight it again.

It weighs less! What's going on?!?

 

[funkstar]

Information in practice is always associated with things that do have mass/energy,

Agreed. That's what "Information is physical" (a Landauer quote) means, imo.

and the information could be considered properties of those things.

If you consider information to be a given property of physical systems, then you have the unpleasant problem that we may have different interpretation functions for the information content of a physical system.

If you start from the information instead, then it turns out that regardless of the physical system embodying the information, physics does have something to say, e.g. with Landauer's principle.

 

[funkstar]

You know about the principle of universal computation then?

I'm not familiar with this. Do you have a reference?

What's the inverse of computation? Reversible computation of course.

In what sense is reversible computation "inverse" to ("normal"?) computation?

 

[Pete]

If you consider information to be a given property of physical systems, then you have the unpleasant problem that we may have different interpretation functions for the information content of a physical system.

If you start from the information instead, then it turns out that regardless of the physical system embodying the information, physics does have something to say, e.g. with Landauer's principle.

Interesting stuff. Thanks funkstar.

 

[funkstar]

Hi Vkothii,

A computer is a physical machine and any computation performed by such a machine is in essence a physical process. This is a simple factual statement but it has a profound consequence. It can be logically argued from this premise that:

* the laws of [physical] computation depend on the physical laws obeyed by the computer machine under consideration, and,

* there are no absolute laws of computation valid for all computational machines

Actually, I've stumbled across the lecture notes (or was it slides?), and the part in red caught my eye. Intuitively, I couldn't disagree more. I don't remember there being any argument for this part. If there is, and you have the notes handy, could you summarize it?

Because it seems that he's essentially disagreeing with the Church-Turing thesis, but the statement has much stronger consequences: For instance, if there are no theorems which holds for all computational machines, then there is a machine to solve the halting problem (in finite time.) That's a rather radical statement.

 

[Vkothii]

I never said or thought that thoughts are not real.

Ok, but you did answer my earlier question, in a way that suggested (to me) that you did think that.

Can you, even try to name anything that is not physical?

Thoughts. Concepts. Ideas. Information. Knowledge.

Take a blank punch card. Weigh it.
Now store a bunch of data on it, and weight it again.

It weighs less! What's going on?!?

What's 'going on' is a difference in mass.
Is the information now punched into the cards, the equivalent of a hole, or the equivalent of the bit of card that used to be where the hole is?

Information, ipso facto, is mass/energy (as its physical representation, which has both).
There is the physical entropy of the representation, and the logical entropy of the representation.
A representation doesn't represent anything, unless someone interprets it (with a brain that uses energy).

Either the bits of card in the bin of the card-punch, are the representation, or the holes in the cards are the representation, which is it? What's the logical difference?

 

[Vkothii]

In what sense is reversible computation "inverse" to ("normal"?) computation?

Well, it depends on your definition of reversible or invertible.
A NOT gate inverts a signal one-dimensionally. If you 'keep' the input there are now two 'states', each the inverse of the other.
The computation is time-symmetrical - forwards gives an inversion, as does backwards, and you can keep both the results around as long as they don't dissipate.

Reversing or inverting a 'computation' are kind of the same thing, aren't they?
The 'principle' of UC is the Universal Turing Machine. But we can't prove it exists, or something.

 

[funkstar]

Well, it depends on your definition of reversible or invertible.
A NOT gate inverts a signal one-dimensionally. If you 'keep' the input there are now two 'states', each the inverse of the other.

But there's no reason to keep the input since a NOT is logically reversible?

The computation is time-symmetrical - forwards gives an inversion, as does backwards, and you can keep both the results around as long as they don't dissipate.

Reversing or inverting a 'computation' are kind of the same thing, aren't they?

I'm still not sure I follow...

 

[funkstar]

The 'principle' of UC is the Universal Turing Machine. But we can't prove it exists, or something.

Again, "exists" in what sense? We're certainly able to define universal TMs, but I agree that no computational machine exists that computes a universal TM...

 

[Vkothii]

But there's no reason to keep the input since a NOT is logically reversible?

What if you keep it anyway?
Does it make any difference? Bearing in mind the 'input' of the reverted computation, is the 'output' of the non-reverted one?

You could say it varies, or is contravariant, wrt to linear transforms in the time-domain?

With the evolution thing, a computation evolves information - after you invert a signal, there it is, you have the information - exactly 1 bit of it. You can hand it on to another computation or hang on to it (or copy it). Quantum information is somewhat different, but only the computation part, and how it has to be reversible, but allow for actual entropy.
I.e. it has to be physically reversible, without actually logically reversing the computation (the direction, or parity of the channel). A NOT gate fundamentally demonstrates the idea of a 'flow' of something to somewhere. A commutation, or translocation of a signal.

I suppose it's the way copy or fan-out happens in a quantum sense, that is kind of the guts of QIS, or that encapsulates what an event is.

 

[Enmos]

Take a blank punch card. Weigh it.
Now store a bunch of data on it, and weight it again.

It weighs less! What's going on?!?

Ok, that's a much better example

 

[Vkothii]

Er, it's a much better example of encoding a message - encoding (a signal, modulating something by "doing work"; transforming; computing).

You compute a stack of physical cards, a representation with a physical extent that has had work done on it - thermodynamic energy $$ H \,= \Delta S\,$$has produced information $$\, I_H \,= I_{\Delta S} $$

This is the equivalent formula to use for moving a bunch of rocks around to make some sort of code. Shannon's formula.

 

[Enmos]

Er, it's a much better example of encoding a message - encoding (a signal, modulating something by "doing work"; transforming; computing).

You compute a stack of physical cards, a representation with a physical extent that has had work done on it - thermodynamic energy $$ H \,= \Delta S\,$$has produced information $$\, I_H \,= I_{\Delta S} $$

This is the equivalent formula to use for moving a bunch of rocks around to make some sort of code. Shannon's formula.

Yea, so it's the organization of the physical stuff that carries the information, not the physical stuff itself :shrug:

 

[Vkothii]

Yes, I know. I've never said that the information is the stuff, it needs to be the stuff, see?
It depends how you read "information is physical", because it means: "information must necessarily always be physical". Information can be a drop of rain falling out of the sky, or a piece of paper blowing along the ground. Anything can be information.

Since we're computers too, we filter most of the incoming messages, or disregard their significance.
Nonetheless, I guarantee if you live in a city, you must recall seeing paper blowing along (or stationary somewhere - in a "steady state" as it were, in terms of bits of paper). So you know it's information - why would you be able to recall instances of it otherwise.

 

[Vkothii]

Polaroid is made from polyvinyl alcohol (PVA) plastic with an iodine doping. Stretching of the sheet during manufacture ensures that the PVA chains are aligned in one particular direction. Electrons from the iodine dopant are able to travel along the chains, ensuring that light polarized parallel to the chains is absorbed by the sheet; light polarized perpendicularly to the chains is transmitted.

So getting back to the question of what does a polarizing filter do to 'unpolarized' light?

Why do polaroid films or lenses reduce glare? what's the "process"?
Apparently it's to do with wavelengths, and waveguides. What sort of 'encoding' occurs, and what does the work?

You have to talk about a wavefunction, remember. Although photons are polarizable in any given axis of rotation, only one axis matters when it comes to the information content.
What is the information content? A wavefunction is a mathematical representation that says a photon has two possible plane-polarized directions, it has to 'polarize itself' along one of two axes perpendicular to the axis of propagation. Or it has to propagate in one direction, so there are two left to choose a polarity along.

In other words: $$ |\psi\rangle\,= \alpha|\uparrow\rangle \,+\, \beta|\rightarrow\rangle$$

The filter absorbs or reflects photons with a horizontal component and transmits those with a vertical component, say. So does the filter select photons with a certain polarity, or does it 'process' photons by giving 1/2 a vertical polarity, and 1/2 a horizontal polarity?

 

[Vkothii]

Apparently, no-one knows the answer?
Do we compute anything by being 'aware' of events and objects, like pieces of paper, or apples on trees that 'fall', so now they're on the ground?

That means - are we biological computers? If we are how do we compute something like polarization of light? Or does the light 'do' the computing, or do a pair of polaroid filters between our eyes and the 'photons', channel something?

Entropy sure explains a lot of things. The entropy of quantum information is just a little trickier, but as you can see, putting on a pair of sunglasses 'encodes' quantum information - remove the specs, and you're back in 'classical' reality.

 

[Vkothii]

"..the spin of a particle classifies how it transforms under the little group, the subgroup of the Lorentz group that preserves momentum.
For massless particles, there is no rest frame. The finite-dimensional unitary representations of the little group [are representations] in two dimensions, the rotations about the axis determined by the momentum. [F]or a photon, this corresponds to the familiar property of classical light - the waves are polarized transverse to the direction of propagation.

Under a rotation about the axis of propagation, the two linear polarization states (|x> and |y>), transform as

$$\;\;\;\;\; |x\rangle\, \rightarrow\, cos \theta |x\rangle\, +\, sin\theta|y\rangle $$

$$\;\;\;\;\; |y\rangle\, \rightarrow\, cos \theta |y\rangle\, -\, sin\theta|x\rangle $$"

So, who would like to be the smart-arse who substitutes these into the above representation, so we transform the 'computational basis' into alpha and beta components of/in the 'standard basis'?

P.S. You should get an answer that includes something like this:

$$ \;\;\;\;\; |x\rangle \,\rightarrow\, e^{i\omega / 2}|x\rangle $$

$$\;\;\;\;\; |y\rangle \,\rightarrow\, e^{-i\omega / 2} |y\rangle $$

Since you have to include your retinal cells as another 'device', in the equation,

 

[Vkothii]

While you're thinking on that, think on this lot:
Photons are the quantum of the electromagnetic field. That means they are the field, in the same sense that waves on a surface, or beneath one, are the field of that surface or volume. The EM field is perturbed by objects that have spin and charge, and mass. Photons are 'spinless', have zero charge and no rest mass, because they are simply a perturbation, where mass is concerned.

But of course this view is interchangeable with the view that mass is the perturbation that photons make, but that would be then in the same sense that waves make boats, by 'sending boats to each other', rather than boats sending waves to each other. Both views are an abstraction, in terms of the information content involved and we usually use the latter version.
So that also means that everything is an equivalent amount of photon information, or "the amount of information in something with mass is the equivalent of the amount of information in photons in the same mass".

Which we've heard before somewhere...?