Physical question
Is there an analogy, and is there a difference, between the ability to perform a task based on stored information, that can be called knowledge, and the ability to perform work based on stored work, that can be called energy?
Comments (44)
What's interesting are the differences.
Energy is conserved, knowledge isn't. I don't deplete my knowledge by using it, nor does my acquisition of it mean that someone else has less of it.
One way to look at the progress of humanity is to see it as the progressive substitution of personal energy by knowledge, which allows us to use other sources of energy than our own, or to accomplish our ends using less energy.
The non-conservation of knowledge has led to a kind of knowledge-communism, making the intellectual achievements of mankind available for very small cost to everyone, a step toward Marx's Kingdom of Freedom.
Yes, indeed. There needs to be compensation for good work, in order for the good work to continue. If I were a billionaire, I'd fund a few 'Open Source Villages' around the world, and provide an apartment, free meals, and a small stipend, for provenly-competent programmers to work on Open Source programs.
Sometimes these people make some money by proving support at a price -- that's how MySQL works, before they sold themselves to Oracle. And people like me, with, in retirement, more income than expenses, can indulge our sense of social responsbility for good things like Video Download Helper, Wikipedia, Calibre, Phet ... by regular somall donations or one-off larger donations.
I would love to see a system whereby after, say, ten years, an author's ebook could be put into a system making it downloadable for, say, a dollar. I suspect that they would get more income from that than from the high prices for eBooks Amazon charges, where they then have to compete with bit-torrents which make the books available for free. I'll bet a lot of people who use bit-torrents feel a bit guilty about the authors' getting nothing ... but not guilty enough to pay Amazon and others' high prices for a digital product.
Obviously there are differences, but there is the analogy also and it's a nice one.
But one needs the other to exist. !!
All knowledge is experiential.
Experience is a by-product of life.
Life requires energy.
Therefore to access knowledge requires work done, and the conversion of stored energy.
Conversion of stored energy is necessary for cultivation of knowledge. But knowledge is not the necessary outcome of converting stored energy.
lol
are you stoned right now?
Obviously we can, for machines built by us. But what about the objects of nature?
Can we use the concept of information for non-human living creatures? It seems to me
we can ... certainly for higher animals. An animal that smells food and moves toward it ... can
we say that this animal is acting on stored information, which associates that particular smell, with
satisfying its hunger? Or should we say that all behavior except human behavior is just conditioned
reflexes? That seems wrong to me.
What about, say, insects? Are bees who do the 'waggle-dance' imparting 'information' about the location of food source to other bees? That seems correct to me.
How far down can we go, before we really are just dealing with conditioned reflexes? Or have I made some sort of category mistake here?
Not just the validity but the very meaning of your view seems dubious to me, regardless what you are or aren't smoking.
"Can we use the concept 'information' to describe the behavior of non-living matter?"
The information that non-living matter have is hard to explain. For example, do you say a planet has information on how to orbit its star? Is the planet the possessor of that information? Which then uses that information to do the orbiting? So, a star knows how to form and how to die? Water knows how to stay liquid and at level? Nitrogen knows how to react with Hydrogen? This type of knowing seems to be much different than the much elaborate information in living things, like a seed knowing how to become a flower, or a cell with DNA knowing how to grow (what size, what cell organelles, how many mitochondria, and so forth). Then there's human information, like knowing where your keys are or the digits of your numberplate. Then the computer information, all the data and instructions and so on. All these information are different. It seems to me the information in/of/with non-living matter isn't the same as these others.
But isn't the only way for information to exist is in binary form? Or am I mistaken here? What number system does nature use? Does it even use numbers? Where is the information for the wall of a beehive to always have six sides or lilies to always have three petals? Is that in the DNA? What is the form of that information? Then again where is the information for the chemicals to have their characteristics? Are both of those really information? Are they both in binary form?
(especially on the question of whether nature uses binary)
Information is an invention of humankind in the sense we are discussing.
It is the name we give to sensual and mental processes.
All living things use information, it is just a question of how it is processed.
Conditioned reflexes are still based on information input.
Touching a hot surface does not result in (at least in the first instance) an internal debate about the intensity of the heat and whether or not to stop touching it. Instead you simply and instinctively retract your hand.
Similarly a bee takes sensual information about the location of pollen and sensually communicates it to fellow bees.
We can indeed 'use concept 'information' to describe the behavior of non-living matter' through machines but only in terms that we can comprehend. Machines are built to interpret data that we think is there, but we can't experience first hand. We then create a way to see what we think we will see.
Why?
So smoking or not (in this case not ) my point is invalid....for what reason
Not just that.
Nature uses the base-one system. All else is the work of man.
I don't think the ocncept of information is useful when describing, say, planets in their orbits. There is no choice involved there. And at the sub-atomic level, things are just random
But at an intermediate level, we have things like a robot vacuum cleaner -- the sort that wanders around your home, avoiding falling down stairs, 'remembering' where obstacles are -- based on previously bumping into them and then storing this information -- where the obstacle was, based on some simple co ordinate system -- and choosing its path using this stored information. (All deterministic of course.)
Another thought: the distinction between 'information' and 'knowledge'. I think these words should apply to somewhat different concepts: 'information' is the reduction of uncertainty. One can have 'information ;which is not yet turned into 'knowledge'. I don't want to quibble about the meaning of words, which don't have inherent meaning anyway. But it seems that these words refer to different although overlapping concepts. Knowledge is brought into being when patterns are spotted in information, and generalizations are made on that basis.
Artificial intelligence from a knowledge/intelligence perspective is not that indifferent to its living counterpart.
But in terms of the subject at hand to same principles apply.
All mediums that contain intelligence, knowledge or otherwise consume energy. We as organisms consume calories for example, and a computational server that contains and artificially intelligent algorithm consumes watts.
I am not (here) interested in the difference between organic life and intelligence. Just the idea between a 'difference [in] between the ability to perform a task based on stored information, that can be called knowledge, and the ability to perform work based on stored work, that can be called energy?'
"Knowledge is brought into being when patterns are spotted in information, and generalizations are made on that basis. "
So could you also say, knowledge is a distinctively human thing, while information is everywhere? But then, does a cheetah knowing how to hunt constitute as knowledge or information? Does a seed of a plant have knowledge or information about how to grow to become a certain type of tree? And is it knowledge or information when hydrogen molecules know that they're lighter than oxygen molecules?
In case of humans. Are these knowledge or information? a) I know how to knit. b) I know the height of Mount Everest. c) I can or I know how to digest complex foods.
Before you where told or saw it, did you know there even was a Mt. Everest, let alone how tall it was.
If you do know how, tell me how to digest complex foods, because I don't know. However even though I don't know, I can also.
You already know it, just ask your gall bladder!
I never said knowledge 'is a distinctively human thing' I said information in the context of 'data', recorded data, recorded by humans.
Knowledge is teachable. Man teaches man to knit. Cheetah teaches cubs how to hunt.
A seed does not use knowledge to grow into a tree anymore than 2 gametes do to become a person. Or a cheetah.
Or Mentos do when reacting with cola.
Hydrogen is lighter than oxygen because be observe a heavy particle and said oxygen. We observed a less heavy particle and said hydrogen. We noticed hydrogen was less heavy than everything else we has observed. Neither one atom nor another knows it is heavy or less heavy, hydrogen or oxygen. It is data we have compiled from knowledge about our reality.
I am not tying to be a jerk however an really enjoy discourse and being challenged.
As for 'knowledge' and 'information' -- maybe the old distinction between 'knowing that' (which involves information) vs 'knowing how' (which may involve something else), will be useful.
That is, there are some systems whose behavior is deterministic, predictable: systems whose motions can be described using Newton's Laws of Motion, his theory of gravity, and a little mathematics. Extend that to electromagnetism, modify Newton by Einstein, and you've still get predictability.
Leave quantum indeterminacy and entanglement aside for the moment. (Or for the time when someone smarter than me can discuss it.)
We are left with (some?) living matter. At some point in the evolutionary change, we get to living matter whose behavior is not always predictable. It's at this point that I think the word 'consciousness' becomes useful -- or unavoidable -- and then the actual structure of consciousness needs to be talked about.
I would suggest that conscious systems have a store of information, and also some -- algorithms? heuristics? some word we don't have yet? -- for processing this information along with sense-data from the outside world, which guides their behavior.
I would also claim that robot vacuum cleaners mimic this, but that it's not necessarily the case that the .... don't have a word here .. 'objects?' with which they process their information -- their stored map of the area they are cleaning, the built-in rules for avoiding falling down stairs, what to do when they detect a large quantity of liquid on the floor -- are the same as ours. Our thinking is not mainly algorithmic.
I could be wrong here, and there are certainly people who claim there is no dividing line between the reflex-responses of a parameceum and the reflex-responses of Doug1943 trying to choose a word for this post. And there people who claim that there is no difference in principle between Doug1943 and a robot vacuum cleaner. In their view, parameceum, Doug, and the vacuum cleaner are just complexes of matter, following physical laws -- if there is any non-determinism, it's quantum randomness. There is no 'free will'.
But Dr Johnson and I say, "Sir, we know our will is free, and there's an end on't." and "All theory is against the freedom of the will; all experience for it."
To sum up: the problems of 'information' and 'knowledge' are tied up with the problem of 'consciousness', and we don't really understand the latter.
http://constructortheory.org/portfolio/the-philosophy-of-constructor-theory/
and more specifically, this:
http://constructortheory.org/portfolio/the-constructor-theory-of-information/
Check this out and tell me what you think.
The same guy, David Deutsch, came up with the following distinction: knowledge is information with causal power:
https://twitter.com/DavidDeutschOxf/status/1055180893483081729
That is his current working definition of knowledge, before that he used to have these:
'Information that is the same across many universes'
'Useful information'
'Information which, once physically instantiated, tends to cause itself to remain so.'
'Information that can program a programmable constructor'
So, he must have thought a lot about knowledge throughout his life. Here is his TED interview in which he mentions that definition:
https://www.stitcher.com/podcast/the-ted-interview/e/56853300
Good point, I can't explain exactly how I walk, even though it is a voluntary conscious activity, unlike digestion. I can say that I move my legs in front of me, alternately, but this doesn't really explain much, or even if it does, I can't explain how I manage to move my leg, or maintain the balance when I am standing, although I can do it.
Exactly, there always must be some material substrate that carries or stores information, but the information itself transcends it. Think of the flow of some idea from one brain to the other, and all possible routes it may follow, and all possible converters on its path.
I managed to exchange a few e-mails with him, sent him a few of my essays and he responded. I respect him also, but I am a little bit more ambitious, trying to understand his constructor theory as much as possible, having a background that I have, and even to question some of his views.
Anyway, the point about digesting complex foods was good, by both Frotunes and 420mindfulness, as this is exactly the distinction between being able (=know how?) to do something, and being able to explain how you do it, ie being able to pass that knowledge. So, there are two interesting cases (at least to me) in which there is know how, not followed by explain how in satisfactory way. First one is about creating the artificial ribosome, that is about grasping the knowledge from molecules, because the best way to show that you know how something works (protein synthesis in this example) is to produce it by your self in a lab. This is a great success because these molecules kept that secret from us for a long time, not explaining much to us by themselves. The other is the case of Alpha Zero, in which superior knowledge of playing various strategic games is created automatically by an algorithm, and stored in a an artificial neural network. However, grasping that knowledge, is not easy for human players, as playing power is not followed by explanatory power. So, even if Alpha Zero can destroy you 1000 times out of 1000 games played, it cannot teach you how to become a better player, the way human coach can. So, another type of software, that interprets that knowledge in humanly understandable way is needed.
Try this:
[url=https://epdf.pub/philosophy-of-information-handbook-of-the-philosophy-of-science.html]Handbook of the Philosophy of Science
Volume 8: Philosophy of Information edited by Pieter Adriaans and Johan van Benthem
General Editors: Dov Gabbay, Paul Thagard and John Woods[/url]
However, if that person is able to write a computer program that is able to produce at least 9 very interesting pieces out of 10, without much input given to it by a skilled person, that would show another level of understanding of the art of musical composition (that would also show a computer programming capability as a bonus). The question is however, is something like that possible at all?
Although, there is a problem of objective evaluation of how pleasing certain music is, my subjective judgment about computer generated music is that I am not impressed with what I heard so far, although I know for a fact that much of what I heard recently, and was pleased with, was composed with the aid of computers. As I am not musically educated, although I listen to music all my life, and have a developed musical interest, I am not clear is the art of musical composition objectively explainable and describable by a computer program? I am divided in half with respect to that question.
https://www.youtube.com/watch?v=Y8UawLT4it0
that is Alpha Zero composing its own music, after just hearing a lot of human composed beautiful music, using the same algorithm that was capable of learning chess, shogi and go, but I am not entirely sure if that really answers the question. At least what I could hear from that link, it doesn't.
Besides the question of comparison of these powers:
explanatory power (ability to cause understanding in other intelligent agents through communication)
descriptive power (how exactly is it different from explanatory power?)
learning power (ability to gain knowledge)
memory power (ability to store information)
constructive power (ability to perform a task based on stored information or received via some communication channel)
working power (based on energy)
here were tackled some questions I thought a lot about by myself too, but didn't manage to comment on yet:
Does information exist in a non-animate world, besides the artificial devices in which it is implemented by humans?
Is knowledge and free will and intellect and cognitive power only a human attribute?
If some animals may process information only through reflex-responses, where exactly is a dividing line between these and those who can do more than just that??
Is anthropocentrism luring as again here to think we are something special, without much justification?
What about plants in the same context?
Determinism in a macro world vs quantum indeterminacy, predictability of living entities and their decisions based on determinism or randomness, is there a link between these facts?