[varsigma (OP)]

On the subject of containers and information.

What kind of container is needed for information? Do you agree with the Wikipedia article that says information is an abstract concept?

Here's a question: in the military, information about the enemy is relevant during wartime, actually it's relevant in peacetime too.
So if some officer decides the best way to get a dispatch to another friendly base is to print it, put the printout in a satchel, then order someone to deliver it personally, where is the information?

That's a trick question, information is everywhere; but where is the information which is relevant to the officer or to the enemy?

When I was enrolled in a course on finite-state automata and formal languages, I recall saying something about information to the lecturer, which was, we decide when information is relevant, he agreed. Information can still be meaningful but be irrelevant. That dispatch will be relevant, but not after the war is over. For instance.

Maybe I'm just listing some attributes of information, whatever it is. Or maybe I'm using the notions of relevance and meaning to put a fence around it.

[alancalverd]

Wikipedia puts it neatly:

Abstractly, information can be thought of as the resolution of uncertainty.

When the war is over, (or more simply when the coin has landed) there is no uncertainty of the enemy's position and capability. General Schwarzkopf was asked, at the outset of the first Iraq war, "what shall we tell the Press?" to which he replied "Right now, tell them nothing. When it's over, tell them who won." Meanwhile, the object of military intelligence is to minimise uncertainty, so in the case cited, it is the words in transit.

Shannon's mathematical formalism of information theory deals with the number of bits (however represented - notches on a stick, pulses on a telephone line, whatever...) needed to reduce the receiver's entropy to a level, (a) such that the receiver is confident that his interpretation coincides with the transmitter's intention and (b) in the presence of random noise.

A lot depends on prior mutual understanding. "Turn left heading wun fife zero  and descend flight level too tree fife" is absolutely explicit and can be heard through a lot of noise, but it presumes that the pilot has a working magnetic compass and an altimeter set to standard pressure, knows left from right, and indeed how to turn and descend an aircraft. The problem with the Hollywood emergency talkdown scenario is that the recovering alcoholic who has just occupied the right-hand seat after the captain has died, doesn't know which dial is which, or what knob makes the plane go up and down, so the instruction hasn't altered his entropy very much. By Shannon's formalism, information content is not determined solely by the transmitter!

[varsigma (OP)]

I'm not sure if I have an example of information which is relevant but its meaning isn't certain.

In physics you manipulate symbols using mathematical logic, you do experiments which are effectively information-gathering exercises.

So an experiment to determine Newton's constant: you have a context, perhaps weights attached to vertically suspended springs. You measure "displacements", you also need to know some other stuff like the mass of the earth, and its radius to calculate a value for G.

But what is it? It is relevant to the force between two masses separated by a distance, but what does it mean? Does it even qualify as information, or is it something else?

[geordief]

I never really appreciated  that information  could be a learning discipline  in itself.

Suppose a sentient being  received  some sensory input ,is there  a maximum amount of physical sensory inputs it must receive so as to output  something like a work of art(maybe extremely primitive)?

Is there a correlation  between the physical input and the mental output or can a minimal physical  input  produce an unrelated large  mental (informational?) output?

[varsigma (OP)]

I never really appreciated  that information  could be a learning discipline  in itself.

I took a course in communications, and I was surprised that information has entropy. At the time I figured it was something somebody borrowed from "real" physics. But what did I know?

[varsigma (OP)]

Suppose a sentient being  received  some sensory input ,is there  a maximum amount of physical sensory inputs it must receive so as to output  something like a work of art(maybe extremely primitive)?

I'd try asking ChatGPT about it. What would you need to tell it, in descriptive terms, so it outputs the required work?

Is there a correlation  between the physical input and the mental output or can a minimal physical  input  produce an unrelated large  mental (informational?) output?

I'd say that depends on what the sentient being knows already.

[geordief]

I'd say that depends on what the sentient being knows already

Yes, I thought that too.The brain creates it's own inputs.Any external input goes through a huge amount of processing  before anything like an output can be  observed.

In fact I don't know if it is possible to correlate an individual input with an individual  output.

[alancalverd]

I'd try asking ChatGPT about it. What would you need to tell it, in descriptive terms, so it outputs the required work?

"Required" disqualifies it from being art. The more closely the customer specifies the end product, the more the process becomes engineering rather than art.

Going back to my air traffic example, "left135 FL235" is a precise specification to line up with the runway and avoid other traffic, to be carried out to the letter and not surprise anyone. "Amaze the crowd with some freeform aerobatics" is a request for an artistic display of the same skills, though the sensible controller still requires the performance to be contained within an allocated box of sky.

[varsigma (OP)]

I'm trying to understand a question that's been posed, about good ol' information.

Here it is: can you put information in a bottle by itself, and detect it?

There is no explanation of what information "by itself" is. My answer would be the bottle is information, so it already has information in it, namely that it is a bottle, a particular arrangement of material which--informs you of its functionality. The shape (a pattern) tells you you can put stuff in it. It says what "bottle" means.

Information by itself is an idea of the nonexistence of information, I suppose. What's your take on this somewhat psychiatric ward question? It suggests a serious misunderstanding of what information actually is. A bottle by itself with no humans around to detect it would be what? Undetected information? Is a mind required here?

[alancalverd]

The cartoonist's desert island "message in a bottle" contains information about the sender's location, which significantly reduces the finder's uncertainly as to the whereabouts of Robinson Crusoe, provided he understands latitude and longitude. If it also contains a date, it will give the receiver a good idea of whether it is worth going to rescue the sender. 

Slightly more practical, the Dead Sea Scrolls (found in jars) tell us a lot about the siege of Masada that we didn't know previously, but still follow Shannon's requirement of prior mutual  understanding, in this case how to read ancient Hebrew.

More abstractly (assuming this is what the questioner thought he was asking), information requires a carrier and a terminator. A bottle serves as both. In the air traffic case, the terminator is implicit because a real pilot (not the Hollywood drunken hero) expects to receive (and acknowledge) two 3-digit numbers in a specific order.

There are classic tales of unterminated  data streams such as "send me ?500 or else...." which raises the temperature (entropy) of a relationship if it is not terminated by ".... collect the goods yourself for ?400". So it's a good idea to put a cork in the figurative bottle.

[varsigma (OP)]

More abstractly (assuming this is what the questioner thought he was asking), information requires a carrier and a terminator. A bottle serves as both.

Actually I thought the questioner believed their question was a way to show that information is a concept, it's not something you can physically put in a bottle. This person is either a bit crazy or doesn't understand what information is.

Information by itself is a phrase that contradicts the fact that it needs a carrier, as you say. And if I start with a lump of ordinary glass and end with a bottle, the information that it is a bottle is self-evident.
So I have put the information "this is a bottle made of glass" into the bottle. I think this questioner would reject my "hypothesis", given they style themselves as a philosopher,

Can I put information in a bottle by itself and detect it? Can I detect a bottle if the bottle is by itself? Where is a bottle when it's by itself? Yikes.
But I can keep asking questions about "the information" to try to lower my uncertainty about what the questioner is asking. Heh.

[varsigma (OP)]

Of course /slaps_forehead.

Shannon explains the question this philosopher posed, and its uncertainty. It's the kind of question a philosopher might believe is an important examination of the nature of information--they presume the answer is no. (Who cares)

But the question contains information and all information has an entropy. The phrase "by itself" can be queried. Is the question "by itself"? no it isn't. Is any part of the question "by itself", evidently not.
Nothing in the universe, apparently, is by itself, everything is connected.

Can I put a label on a glass bottle that says "this is a bottle made of glass", without changing any meaningful information-about-bottles? I'm implicitly communicating with someone or some thing that can decode what's on the label. Without the label, the shape and the glass communicate some information to an observer.
So the question turns into quite a good example of what Shannon was talking about, back when.

Easy peasy. And yes I think philosophy sucks too.

[varsigma (OP)]

Now I can crack some IT jokes.

A pair of informations go into a bar. The bartender asks, "You're by yourself, you two?"

[alancalverd]

I think this questioner would reject my "hypothesis", given they style themselves as a philosopher,

You have correctly defined a philosopher as a person who makes a living by telling you that you don't understand what you just said. A narcissist with no redeeming features.

[Zer0]

what worries you...Masters you.

[varsigma (OP)]

I used to worry about being able to finish coding a program. Nowadays I'm a bit gobsmacked by the myriad misconceptions, in the age of information, of what it is.

Leonard Susskind says information is entanglement. Entanglement is another well-misunderstood term.
If I think about what he's saying, that implies electrons are entangled with protons in ordinary atoms: they are "charge-entangled". Where is the information? How do you measure it.

There's another word that you need a good understanding of the meaning of. The protons and the electrons in orbitals are measuring each other, just not classically.
If you want classical measurements you will need some classical equipment; today you can scan your solid state of choice with a scanning-tunneling electron  microscope. You recover classical information because the measuring device has a much larger phase space than the sample being scanned.

[alancalverd]

The information is the position of an electron. You have no idea where a free electron might be found, but if you know the location of a proton you have significantly reduced your uncertainty.

We use search dogs to solve the inverse problem.

[varsigma (OP)]

The information is the position of an electron. You have no idea where a free electron might be found, but if you know the location of a proton you have significantly reduced your uncertainty.

Electron position has more entropy than proton position because of the mass difference (I guess). But classical measurements sort of smear this out, because classical measurements are always a representative sample.

When you say free electron,  do you just mean not bound to a proton? As in, flowing in a conductor?
Sorry,  I see you were just commenting on the different states, and how a bound electron has less entropy because protons or atomic nuclei, in general, are easier to locate. I'm recovering from a cold so a bit woolly.

[varsigma (OP)]

ChatGPT seems to agree with the, ah, philosophy that information gathering, or signals intelligence, is about recognizing patterns.

These patterns might be seen in the activities of enemy communications, even if what is being communicated isn't known. Information with a high degree of uncertainty because it's encrypted is commonly sent and received in military contexts.

Patterns are physical, information in patterns is necessarily physical too. ChatGPT also agrees that the shape of physical objects is one of these physical patterns.

An ordinary glass wine bottle say, "remembers" its shape. It is a store of information even when its empty.
I can't understand why anyone shouldn't see this obvious thing. But it seems some people just can't.                 

[alancalverd]

ChatGPT doesn't recognise that "pattern" has several meanings. SIGINT may involve recognising the start or end of a message but if the content followed a pattern it wouldn't be information since the receiver already knows it the second time he gets it! If the transmitter is using pseudorandom frequency hopping, you need to know the code of the day so that the receiver can hop in sequence, and remembering the sequence won't be any use to a third party because the seed code will be randomly different tomorrow.

There are classic cases of ENIGMA and TUNNY decodes that depended on human stupidity such as always finishing with HH, sending the same message twice with the same seed, or the interceptor knowing that weather readings are always transmitted on the hour, but I doubt that such patterns occur much these days - everyone has read the book!