< 1 2
 
   
 

Complexity and Design - Two Arguments

 
Poldano
 
Avatar
 
 
Poldano
Total Posts:  3152
Joined  26-01-2010
 
 
 
10 July 2017 19:42
 
no_profundia - 08 July 2017 09:12 AM

...


NL, I imagine Poldano could give a better answer to this question but my understanding is that entropy is a statistical property of the system as a whole. You won’t necessarily see any difference at the level of the behavior of individual atoms. To use Sean Carroll’s favorite example, if you have a cup of coffee and cream and all the cream is on top and all the coffee is on the bottom that is a low entropy state compared to the case where the coffee and cream are all mixed together because if you added up all the possible ways the atoms could be arranged that would produce the coarse-grained description “all the cream on top” the number would be much smaller than the number of possible ways the atoms could be arranged that would produce the coarse-grained description “all mixed up”.

But entropy is still quite real. I don’t think it is merely a human abstraction. If the universe had not begun in a state of very low entropy stars, planets, life, etc. would never have existed.

Entropy is a mathematical abstraction, but it is an abstraction that was developed specifically to address real phenomena.

Entropy in its thermodynamic usage is a whole-system property. I’m not so sure of entropy in its information-theoretical usage, because that’s defined as a function (logarithm specifically) of the maximum amount of information that a signal can contain.

To clarify my position on this, for what it’s worth, anytime we can talk about something we are dealing with abstractions, because language as we know it requires predication, which is a form of abstraction. Generally, this only matters when one side in an argument asserts that some particular abstraction is real, and uses that assertion to deny the validity of another abstraction. Such arguments are invariably bogus, but those who believe in the reality of their own abstractions will not be convinced of the bogosity.

Is that Zen enough?  wink

[ Edited: 10 July 2017 20:03 by Poldano]
 
 
Poldano
 
Avatar
 
 
Poldano
Total Posts:  3152
Joined  26-01-2010
 
 
 
10 July 2017 19:53
 
Phate - 07 July 2017 08:55 AM

Funny, I was just thinking about this the other night as well, but from a bit of a different angle. Also, please note this is not my personal belief but just some random thoughts for discussion.


Could it be stated that a Universe, taken as a whole, needs to evolve toward complexity?

The evidence for this is simply that 1) We are complex. 2) We are a product of our universe. 3) We have no evidence of other universes that have not evolved towards complexity.

In this sense, the universe was not designed to create us, but rather that it has an instinct (for lack of a better word) to move in a direction that given enough time will lead to complexity, and for the purpose of this discussion I am suggesting that we are the most complex entity in the universe. This aligns more with the bottom up approach.

An analogy to this would be modern machine learning programs. Developers don’t give fledgling AI everything they need to figure out how to play Go. They give the program basic rules that allow it to begin learning what colour is, what shapes are, what the cause and effect of moving pieces is, etc. etc. (yes I stole this part from the podcast) and then they give it a goal (win the game) and let it learn and self organise until it solves the problem.

As such, all a Universe needs is the goal (this underlying instinct) to progress towards complexity. Natural laws, physical matter, etc, etc are just the system self organising itself towards the goal.


Okay, so probably a lot of wrong-turns above but since I have gone this far might as well drive right off the cliff…


To bring this back to religion, wherever the universe came from (god, infinite ether, the division of nothing into matter and anti matter, etc) one could posit that it would have 1) this instinct for complexity, 2) the potential within it to be as complex as we are…and could be in the future.

Essentially, we were not so much designed, as born. Further, we share the same DNA (complexity) as our parent and share the same instincts (evolve towards higher complexity).

Religions (along with culture, politics, etc) therefore is simply our manifestation of this instinct. To self-organise ourselves in a way such that we are able to achieve the goal…further complexity. And this goal is meaningful because the Universe needs to go in this direction (as we have no evidence to the contrary).

———
I can think of several flaws in the above but curious to get the reaction of others first.

I wouldn’t frame it exactly as you do. I think that high complexity needs high entropy, and I strongly suspect that evolution of any kind (including learning or for that memory any phenomena we can refer to as mental) needs increasing entropy.

 
 
no_profundia
 
Avatar
 
 
no_profundia
Total Posts:  372
Joined  14-07-2016
 
 
 
12 July 2017 19:17
 
Poldano - 10 July 2017 07:42 PM
no_profundia - 08 July 2017 09:12 AM

...


NL, I imagine Poldano could give a better answer to this question but my understanding is that entropy is a statistical property of the system as a whole. You won’t necessarily see any difference at the level of the behavior of individual atoms. To use Sean Carroll’s favorite example, if you have a cup of coffee and cream and all the cream is on top and all the coffee is on the bottom that is a low entropy state compared to the case where the coffee and cream are all mixed together because if you added up all the possible ways the atoms could be arranged that would produce the coarse-grained description “all the cream on top” the number would be much smaller than the number of possible ways the atoms could be arranged that would produce the coarse-grained description “all mixed up”.

But entropy is still quite real. I don’t think it is merely a human abstraction. If the universe had not begun in a state of very low entropy stars, planets, life, etc. would never have existed.

Entropy is a mathematical abstraction, but it is an abstraction that was developed specifically to address real phenomena.

Entropy in its thermodynamic usage is a whole-system property. I’m not so sure of entropy in its information-theoretical usage, because that’s defined as a function (logarithm specifically) of the maximum amount of information that a signal can contain.

To clarify my position on this, for what it’s worth, anytime we can talk about something we are dealing with abstractions, because language as we know it requires predication, which is a form of abstraction. Generally, this only matters when one side in an argument asserts that some particular abstraction is real, and uses that assertion to deny the validity of another abstraction. Such arguments are invariably bogus, but those who believe in the reality of their own abstractions will not be convinced of the bogosity.

Is that Zen enough?  wink

I think I would agree with this as long as we add that our abstractions (at least, our good ones) have some basis in reality. The temperature of a gas, I believe, is based on the average kinetic energy of the atoms that make up the gas. This is an abstraction. It is possible no actual atom has a kinetic energy equal to the average. But it is an abstraction that has some basis in reality. Gases with different temperatures will have different effects on the world.

If we had the capacity to measure the direction every atom in a gas was moving, and assign it a number between 0 and 360, we could determine what the average direction was for the gas as whole, but as far as I know this abstraction would not pick out anything interesting or any “real property” of the gas. So, while I agree that language introduces abstractions that may not be part of our most fundamental description of the world it seems to me there are good and bad ways of carving up the world.

 
 
Poldano
 
Avatar
 
 
Poldano
Total Posts:  3152
Joined  26-01-2010
 
 
 
13 July 2017 20:04
 
no_profundia - 12 July 2017 07:17 PM

...

I think I would agree with this as long as we add that our abstractions (at least, our good ones) have some basis in reality. The temperature of a gas, I believe, is based on the average kinetic energy of the atoms that make up the gas. This is an abstraction. It is possible no actual atom has a kinetic energy equal to the average. But it is an abstraction that has some basis in reality. Gases with different temperatures will have different effects on the world.

If we had the capacity to measure the direction every atom in a gas was moving, and assign it a number between 0 and 360, we could determine what the average direction was for the gas as whole, but as far as I know this abstraction would not pick out anything interesting or any “real property” of the gas. So, while I agree that language introduces abstractions that may not be part of our most fundamental description of the world it seems to me there are good and bad ways of carving up the world.

Our abstractions might have no basis in reality, but they probably do. If they did not, we would probably not have evolved them. To be more precise, for basis in reality, I read utility for interacting with reality.

By the way, the figure is not direction but velocity, which is a combination of direction and magnitude, and the range of the direction is not 360 degrees (or 2*PI) but 720 degrees (or 4*PI, the ratio of the area of a sphere to the square of its radius). I think the direction might be a complex number or some other kind of vector as well, because the area of a sphere is two-dimensional, but I would yield to someone who knew the appropriate math on that point.

 
 
no_profundia
 
Avatar
 
 
no_profundia
Total Posts:  372
Joined  14-07-2016
 
 
 
13 July 2017 21:40
 
Poldano - 13 July 2017 08:04 PM
no_profundia - 12 July 2017 07:17 PM

...

I think I would agree with this as long as we add that our abstractions (at least, our good ones) have some basis in reality. The temperature of a gas, I believe, is based on the average kinetic energy of the atoms that make up the gas. This is an abstraction. It is possible no actual atom has a kinetic energy equal to the average. But it is an abstraction that has some basis in reality. Gases with different temperatures will have different effects on the world.

If we had the capacity to measure the direction every atom in a gas was moving, and assign it a number between 0 and 360, we could determine what the average direction was for the gas as whole, but as far as I know this abstraction would not pick out anything interesting or any “real property” of the gas. So, while I agree that language introduces abstractions that may not be part of our most fundamental description of the world it seems to me there are good and bad ways of carving up the world.

Our abstractions might have no basis in reality, but they probably do. If they did not, we would probably not have evolved them. To be more precise, for basis in reality, I read utility for interacting with reality.

By the way, the figure is not direction but velocity, which is a combination of direction and magnitude, and the range of the direction is not 360 degrees (or 2*PI) but 720 degrees (or 4*PI, the ratio of the area of a sphere to the square of its radius). I think the direction might be a complex number or some other kind of vector as well, because the area of a sphere is two-dimensional, but I would yield to someone who knew the appropriate math on that point.

I think you might be more of a pragmatist than I am. I am not sure the truth of our abstractions is merely a matter of utility. Let me ask you a couple of questions and see if I can get a better idea about your position. The fundamental question I had about your position involved this claim “Generally, this only matters when one side in an argument asserts that some particular abstraction is real, and uses that assertion to deny the validity of another abstraction. Such arguments are invariably bogus.”

Are you saying that all abstractions are equally valid? Is average velocity just as valid a description of reality as temperature or entropy? Is there no difference between those abstractions other than their utility for us?

I don’t think I agree that utility is the fundamental concept that is capable of making sense of our notion of reality. Utility is also an abstraction and it is an abstraction that only makes sense when we are talking about beings that have goals and can find some things useful and other things useless.

But if entropy were not in some sense real, if the universe were not evolving from low entropy to high entropy, there would not be any beings with goals and so there would be no such thing as utility. So I don’t think you can use the concept of utility to explain the reality possessed by a concept like entropy since if entropy did not have some basis in reality there would be no such thing as utility.

I tend to think that there are real patterns in the world and those patterns exist independently from whether we ever notice them, or formulate them, or find them useful. While it is probably true that some arguments are based on people mistakenly clinging to the reality of their own abstractions and denying reality to equally valid abstractions I am not convinced that every abstraction is equally valid.

I’m not sure if that is your position or not but I would be interested to hear your thoughts.

 
 
Poldano
 
Avatar
 
 
Poldano
Total Posts:  3152
Joined  26-01-2010
 
 
 
15 July 2017 03:12
 
no_profundia - 13 July 2017 09:40 PM

...

I think you might be more of a pragmatist than I am. I am not sure the truth of our abstractions is merely a matter of utility. Let me ask you a couple of questions and see if I can get a better idea about your position. The fundamental question I had about your position involved this claim “Generally, this only matters when one side in an argument asserts that some particular abstraction is real, and uses that assertion to deny the validity of another abstraction. Such arguments are invariably bogus.”

Are you saying that all abstractions are equally valid? Is average velocity just as valid a description of reality as temperature or entropy? Is there no difference between those abstractions other than their utility for us?

...

All abstractions are not equally valid. Their validity originates in their utility. Different abstractions have different utilities. It is difficult to compare the utilities of different abstractions because utility is case-dependently variable. Different abstractions can also differ in form, and their utility is often dependent on their form as well as their correspondence to situation. Internal coherence, i.e., whether abstractions make sense logically, is an element of form but not all of it.

no_profundia - 13 July 2017 09:40 PM

...

I don’t think I agree that utility is the fundamental concept that is capable of making sense of our notion of reality. Utility is also an abstraction and it is an abstraction that only makes sense when we are talking about beings that have goals and can find some things useful and other things useless.

...

In my view, abstractions are only artifacts of mind, and mind seems to be a feature of beings that have goals. In that view, utility is an abstraction that is dependent on mind, while mind is a real phenomenon that cannot exist without some real phenomena that correspond to both utility and susceptibility to abstraction. The real correspondents don’t need to be perfectly useful or perfectly susceptible to abstractions, however.

no_profundia - 13 July 2017 09:40 PM

...

But if entropy were not in some sense real, if the universe were not evolving from low entropy to high entropy, there would not be any beings with goals and so there would be no such thing as utility. So I don’t think you can use the concept of utility to explain the reality possessed by a concept like entropy since if entropy did not have some basis in reality there would be no such thing as utility.

...

I don’t believe we can explain reality. We can only explain our perceptions and conceptions of it. Where our perceptions in the form of facts correspond to what our conceptions in the form of theories and hypotheses predict, we tend to conclude that we have an understanding of reality within the domain of those conceptions. Strictly speaking, we are not thereby explaining reality, but explaining the validity of our conceptions of reality.

no_profundia - 13 July 2017 09:40 PM

...

I tend to think that there are real patterns in the world and those patterns exist independently from whether we ever notice them, or formulate them, or find them useful. While it is probably true that some arguments are based on people mistakenly clinging to the reality of their own abstractions and denying reality to equally valid abstractions I am not convinced that every abstraction is equally valid.

...

This is the common scientific viewpoint. I don’t absolutely disagree with it. My opinion is that we cannot distinguish the validity of this view from the validity of view that patterns of perception and conception are mental emergences. I agree that there are correspondences, but assert that those correspondences could be imperfect yet still be sufficiently useful to persist. One of the reasons that I take this position is that perception and conception are subject to the rules of computational theory. The utility of a pattern is not just its conformity to real phenomena, but to the ability of physically-embodied mental apparatus to perform the equivalent of computations using that pattern. This co-dependence basically removes the clear distinction between objective (i.e., pertaining to reality) and subjective (i.e., pertaining to the mind) causation at the conceptual level, and probably at the perceptual level as well.

While we may not be able to determine whether patterns actually exist in reality or are mental phenomena, it generally does not matter, but may matter in some specific circumstances. For some purposes (utility again) we can regard patterns as existing in reality. For other purposes, it is better to regard patterns as existing only in minds. The latter is important in psychotherapy, in learning new and difficult concepts, and in scientific paradigm shifts, among other things. It may be convenient to think of this as “thinking outside the box.”

[ Edited: 15 July 2017 03:20 by Poldano]
 
 
no_profundia
 
Avatar
 
 
no_profundia
Total Posts:  372
Joined  14-07-2016
 
 
 
16 July 2017 16:45
 
Poldano - 15 July 2017 03:12 AM

I don’t believe we can explain reality. We can only explain our perceptions and conceptions of it. Where our perceptions in the form of facts correspond to what our conceptions in the form of theories and hypotheses predict, we tend to conclude that we have an understanding of reality within the domain of those conceptions. Strictly speaking, we are not thereby explaining reality, but explaining the validity of our conceptions of reality.

What I really meant here was, I don’t think our concept of reality can be explained with the concept of utility. It seemed to me you were arguing that the notion we have of reality can be entirely cashed out, so to speak, with the notion of utility. In other words, what we mean when we say something is real is it is useful for us to treat it as real.

So with the concept of utility we have reached the bottom or at least the bottom for us. We can’t probe any deeper. This could be true but I am not convinced it is because I think the concept of utility itself is dependent on other things that we need to treat as real in a sense that cannot be explained by referring them back to the concept of utility. Otherwise, we wind up in a vicious circle.

If we say “we consider entropy real because it is useful for us to do so” but the notion of “useful” depends on the existence of beings that can find things useful and that in turn depends on the “reality” of entropy then I don’t think the “reality” of entropy can be exhausted by the concept of utility.

I am not entirely convinced by my own argument here because I don’t think entropy is probably a part of the most fundamental description of the world. The evolution of the universe could probably be explained without making use of the notion of entropy at all so we could say it is just a useful abstraction.

But it still seems to me it must correspond to something real that is not dependent on the concept of utility for its reality.

Poldano - 15 July 2017 03:12 AM

I agree that there are correspondences, but assert that those correspondences could be imperfect yet still be sufficiently useful to persist.

I agree with this completely. I feel fairly certain that most of the correspondences are very imperfect and yet survive because they provide enough of a grip on the world to be useful. I just am uncomfortable with fully cashing out the correspondence in terms of utility. I want to say there is something real there that we get an (imperfect) grip on that does not depend on our minds or the notion of utility.

Poldano - 15 July 2017 03:12 AM

The utility of a pattern is not just its conformity to real phenomena, but to the ability of physically-embodied mental apparatus to perform the equivalent of computations using that pattern. This co-dependence basically removes the clear distinction between objective (i.e., pertaining to reality) and subjective (i.e., pertaining to the mind) causation at the conceptual level, and probably at the perceptual level as well.

I find these ideas very interesting but we are approaching - and probably over-stepping - the outer reaches of my understanding and what I feel comfortable even attempting to talk about. Would you have any books to recommend by any chance that develop these ideas further?

 
 
Poldano
 
Avatar
 
 
Poldano
Total Posts:  3152
Joined  26-01-2010
 
 
 
17 July 2017 01:22
 
no_profundia - 16 July 2017 04:45 PM
Poldano - 15 July 2017 03:12 AM

I don’t believe we can explain reality. We can only explain our perceptions and conceptions of it. Where our perceptions in the form of facts correspond to what our conceptions in the form of theories and hypotheses predict, we tend to conclude that we have an understanding of reality within the domain of those conceptions. Strictly speaking, we are not thereby explaining reality, but explaining the validity of our conceptions of reality.

What I really meant here was, I don’t think our concept of reality can be explained with the concept of utility. It seemed to me you were arguing that the notion we have of reality can be entirely cashed out, so to speak, with the notion of utility. In other words, what we mean when we say something is real is it is useful for us to treat it as real.

So with the concept of utility we have reached the bottom or at least the bottom for us. We can’t probe any deeper. This could be true but I am not convinced it is because I think the concept of utility itself is dependent on other things that we need to treat as real in a sense that cannot be explained by referring them back to the concept of utility. Otherwise, we wind up in a vicious circle.

If we say “we consider entropy real because it is useful for us to do so” but the notion of “useful” depends on the existence of beings that can find things useful and that in turn depends on the “reality” of entropy then I don’t think the “reality” of entropy can be exhausted by the concept of utility.

I am not entirely convinced by my own argument here because I don’t think entropy is probably a part of the most fundamental description of the world. The evolution of the universe could probably be explained without making use of the notion of entropy at all so we could say it is just a useful abstraction.

But it still seems to me it must correspond to something real that is not dependent on the concept of utility for its reality.

...

It’s terribly easy to get into circular reasoning about this stuff. My own concept of reality is that the term reality is a pure reference. Any time we describe reality or talk about objects or phenomena that are real we are actually referring to a model of reality. The ability of two agents to talk about something coherently is strong evidence that the words used for that something actually refer to something real. If actions are performed such that two agents can observe and agree upon an outcome of the actions, the evidence for reference to something real increases further. However, the descriptions are never completely exhaustive if the reference is part of empirical reality. (Formal deductive systems are different, but that’s too much to go into right now.)

Where utility comes in is that the basic tools we use to reference reality, and the ways we classify real things, are dependent on utility. Reality itself isn’t dependent on utility (it isn’t dependent on anything), but any references to it are dependent upon utility. This applies to tools like mathematics and language as well as to the ways we organize our knowledge of empirical reality. I would even go so far as to say that if some element or aspect of reality is not useful to us in any way (and those concepts indicate that we are already speaking about our model rather than about reality itself), we are unaware that it exists.To be sure, utility is not just pragmatic in a colloquial sense, but can also be helpful with arts and sciences that are sometimes called “pure”, because they have no economic benefit. We humans are addicted to explanations, to trying to make sense of ourselves and our world, even if there is no obvious benefit to it, and my notion of utility applies even there.

Here is a diagram that may help:

          REALITY ==>  UTILITY-DRIVEN INTERPRETATION ==>  MODEL OF REALITY

We cannot really talk about or even perceive the REALITY before the first arrow. We mostly talk about the MODEL OF REALITY after the second arrow. With some difficulty we may be able to talk about UTILITY-DRIVEN INTERPRETATION, but even that is most likely about a model of that process. We can improve UTILITY-DRIVEN INTERPRETATION with a great deal of difficulty but observing differences between what the MODEL OF REALITY predicts and what actually happens. This is a largely stochastic process because we have difficulty considering UTILITY-DRIVEN INTERPRETATION apart from the MODEL OF REALITY it generates. Consequently we must rely on a great deal of guesswork to tentatively replace one set of interpreters with another and guess or observe the outcome in order to make improvements.

I hope this helps to clarify my viewpoint.

 
 
 < 1 2