How Does the World Work?


  • See the About page for a description of the subjects of interest covered in this blog.

Series Indexes

Global Issues Blogroll

Blog powered by Typepad

Comment Policy

  • Comments
    Comments are open and welcome as long as they are not offensive or hateful. Also this site is commercial free so any comments that are offensive or promotional will be removed. Good questions are always welcome!

« | Main | Summer Solstice - 2016 »

May 06, 2016

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

Pedro Lopez

So since this is a formal system, how does it account for the implications of Godel's theorem and of the field of non-Computational Mathematics? In other words the inherent limitations of Formal Systems and systems that are prove-ably intractable to algorithms.

Kevin Baas

How is this different from a universal turing machine?

http://www.dna.caltech.edu/courses/cs129/caltech_restricted/Turing_1936_IBID.pdf

Kevin Baas

..or Mathematica's programming language? http://www.dna.caltech.edu/courses/cs129/caltech_restricted/Turing_1936_IBID.pdf

Kevin Baas

(sorry bad ling .. no edit button...) https://reference.wolfram.com/language/guide/LanguageOverview.html

George Mobus

I should have anticipated something like this. Call it QE Ascending (in hits).

Someone posted this to slashdot (/.) a computer "science", hacking website. The community is huge so as a consequence as of this morning 10:00 PDT I've gotten over 2000 hits! Kevin and Pedro's comments above are mild (and polite) versions of the comments to be found there. If non-computer readers are interested in, shall we say, criticisms, go to: https://developers.slashdot.org/story/16/05/07/187215/researcher-writes-a-machine-language-for-the-universe

As one who has taught computer science students for nearly 25 years I am not at all surprised at some of the blunt if not vitriolic comments found there. The most frequent complaint I hear from students is how they could not understand what they read in the textbook or the exam questions. I have conducted several reading workshops over the years to try and find out why these students (and it is the vast majority of them) were having trouble. It turns out to be simple and somewhat scary. They never learned how to read. For meaning that is.

It is actually even worse than that. A simple question: Why do so many software development projects still end up failing in some ways to meet 'user' requirements? Answer: IT people have difficulty understanding anything that isn't computation, and most of life isn't (in the strict sense - see below). It isn't just reading that is a problem, it is also just not comprehending non-technical language well enough.

The slashdot community is a great example of the kind of siloed disciplinary approach that has been a source of the failures of science and engineering to recognize the systemic nature of the world and the problems that causes. I suspect a large majority of them think that computer science (and related mathematics) is THE only legitimate domain of interest and that all questions eventually come down to computation. Ergo, most of the comments (such as above) are challenges framed in computer science and mathematics concepts.

My brief description of SL was taken (quickly) out of context, namely that I am talking about the language of building models of systems, not explaining, for example, quantum phenomenon (unless, for example, loop quantum gravity (LQG) turns out to be "right" because there is a lot of systemness about that theory!) Therefore, Pedro and Kevin, your questions are misguided in this context. For example, Gödel's incompleteness theorems still (always) apply in the case of pure mathematics, but so what? Most mathematicians and especially the applied version (of which this is an example) carry on somehow. No computer stopped working after Gödel proved his results. Similarly asking about differences in SL versus a UTM misses the point. Kevin, look at: Why Interaction is More Powerful than Algorithms by Peter Wegner for an expansion of the notion of computation beyond Turing. There are things systems do that are not technically Turing computable. Also, Wolfram's "A New Kind of Science" looks an awful lot like an attempt to explain the universe with cellular automata. But he doesn't seem to have made much progress beyond that original idea.

The idea of finding a language which is capable of describing and modeling complex dynamic systems in a way that is comprehensible to both ordinary human beings and computers should actually be acceptable to computer scientists (most of my colleagues don't have any problem with it at all). That I am proposing to develop a language that is fundamental to our minds and translatable into a formal structure should come as great news. Unfortunately the slashdot community (the most vocal ones anyway) read into the blog what they wanted and reacted with derision. Fortunately it won't matter insofar as my work is concerned.

George

Cliff

The slashdot community can be very nasty and immature. It is a reflection on them - not your ideas.

Lucas Jordan

As a regular slashdot reader (and a computer scientist) I can assure you the majority of those comments are missing the point of this work.

I look forward to seeing what you put forward in July.

In all truth, a simple language (with good tools) that can express systems and provide a model to explore them, would put a lot of software people out of work :)

Doug Mayhew

I am impressed by your boldness and top-down method, which I believe is the better way to find out how things work. One of my favorite learning principles is, if you get stuck, it's usually because you are zoomed in too closely to see the all-important context within which a given problem is situated. Walking up and down the ladder of abstraction is good, but sometimes, you have to break new ground because the ladder itself doesn't go high enough to organize a pattern of particulars.

There, in abstractions of abstractions, the air is thin and ideas often vanish without great care and effort. But that's where all the great discoveries happen historically.

Don't worry about scarce resources being civilization's main negative feedback force. It isn't. That's more or less foolish propaganda aimed at destroying creative minds like yours. I would blame oligarchical usury as the main historical challenge facing mankind. Folks in that club spend a lot of time making resources scarce for the rest of us, so they can rule us or use us in one way or another. What they fear most is folks who can see that scarcity is a false construct, whose inversion is faster rates of unfolded human potential. So many existing resources are there because someone discovered them and found ways to use them. We have to generate and act on more powerful and profound ideas and discoveries in order to stay ahead of entropy or those who drag our collective feet as we try to advance.

Historically, the bottleneck to progress is not so much mother nature's reality or scarcity or resources, but bullshit accepted as fact by the people and their politicians.

Man expands his population potential with each new discovery, and with each upward leap to a higher density energy sources. This is even more true of cultural and political innovations which recognize this principle as the primary organizing element. So there's your first recursion. :-) This is why you can't model man like you model rabbits. Animals will never discover Fusion technology, no matter how long you run the simulation. There is a qualitative difference, or unique principle within human minds, related to an ability to engage in highly abstract thought. This unique difference relative to animals is difficult to pin-point, yet efficiently expressed by our rise to a population of billions over our nearest cousins in the animal kingdom, and also by the formation of complex political economies on a planetary scale. Soon we will establish colonies on other worlds. In computer science terms, its like animal minds are NP or NP Complete "level", and human minds are NP Hard "level".

Godel became famous for showing some hard limits to formal systems, and he hung out with Einstein at Princeton. I don't think you should "so what" away Godel, in the sense that the things you are trying to model probably are going to run smack dab into Godel's area of expertise.

It smells like you need to invent a new kind of computer, (likely non-digital or at least non-sequential) before you can model Mind or the Universe. But I do encourage you to investigate such things, and let us hear about it.

Like you say, the field is so subdivided into specialties, few are paying attention to the super-abstractions, and I am willing to bet many important discoveries will come from them. I think perhaps being limited to digital machines as they are, is not the best place to start necessarily, and this is yet another example of specialization potentially suffocating most of the fields computer science and system theory.

Just look at any field, say biology: Mankind is trying to figure out how DNA works using mostly digital methods. We still don't have a clue as to how the whole DNA system is organized, just relatively disconnected islands of knowledge with an expert in a lab coat standing on each one waving to the others.

If need is the mother of invention, then you are in the right place, but you might have to reinvent the computer before going after the things you want to model.

Finding an alternative computational device in itself would be an awesome adventure. Maybe that's how it has to proceed naturally anyway, define the blockade, and then find a way to breach it. I will be following your work, and I wish you luck.

Doug Mayhew

As a side note, I am very interested in interactive computing models, and I found the paper you linked to in your post above on that topic very insightful, esp in regard to interactive models ability to exhibit behaviors beyond the reach of algorithms. Almost sounds like an echo of the power of Plato's Dialectic, "2 way discussion seeking insight", having superior power to penetrate into the nature of things.

https://en.wikipedia.org/wiki/Dialectic

Short of inventing a new type of computer, perhaps you can experiment with the Erlang computer language to form the communications boiler-plate infrastructure to enable many thousands or even millions of isolated interactive instances across a cluster of machines very efficiently. It seems to fit the model of "interactions via messages only" quite well and can be scaled to enormous sizes in a computing cluster. I would like to see what folks in your league of understanding could do with such a beast. Erlang is relatively simple to set up, too. I don't know how it's design would limit your scope of investigations, but the article about interactive computing sounded very Erlang-ish to me. Maybe your language could be made to "ride on top" of Erlang for execution, assuming you need zillions of agents and supervisors, etc, without bogging down the resources too much. Just a thought.

Contributor

Does emphasising computational methods or machines confirm the 'homo callidus' versus 'homo sapiens'idea behind much of this blog's message?

Koen Lefever

George Mobus wrote: "The slashdot community is a great example of the kind of siloed disciplinary approach that has been a source of the failures of science and engineering to recognize the systemic nature of the world and the problems that causes."

Slashdot has many vices, but definitely not that one.

Dismissing the /. community as a "computer "science" (why the quotes?) and IT silo" is far besides reality. Biologists, physicists, mathematicians, philosophers (I'm a philosopher of science and logician myself, active on Slashdot since 1998), historians, sociologists, geographers, linguists, etc... do participate actively in discussions. It is a site where inter-disciplinary dialogue has been going on for almost two decades.

George Mobus

Thanks to Cliff, Lucas, & Doug for supportive comments. Doug, I have your e-mail so will respond via that channel when time permits.
@Contributor,

Since I was corrected once on the spelling (I do not know Latin) it is Homo calidus, not callidus. And I really don't know what you are getting at.

Which brings me to two points re: these comments and those on /. 1) I will not respond to single sentence supposed zingers as these demonstrate considerable lack of critical thinking. You need to explain your criticism not just blast a one-liner thinking that is sufficient.
2) The vast majority of the comments on /. and those at the start of this thread indicate that readers (or should I say perusers) failed to grasp that this language is about modelling not about a deep description of everything. I used the term 'universal' only to emphasize that the language should be able to model any identifiable system (a thing that embodies the principles of systemness). Models are always necessarily reduced and compact descriptions of that which they model. As such they are not expected to explain absolutely everything in some reductionist view (such as a presumed GUT). Rather they are practical ways to anticipate the future through a partial explanation of how a system works. However, the more detail they can provide, the more we can say we 'understand' that which is modeled. My claim is (or will be soon) that SL provides a much more 'complete' way to describe systems. It is NOT a way to describe the universe down to the Planck scale.

@Koen,

Granted you may be right. I've somehow missed those conversations. Perhaps I should have been clearer in specifying that portion of the /. community responding to my blog (or rather to the article in /.) They are clearly of the IT persuasion and many have enough knowledge of CS to make specific barbs.

My question is: What major new ways of addressing the interesting interdisciplinary problems have emerged from said conversations? If any, perhaps I should join!

dk

The universe itself is constrained by limits similar to the Godelean incompleteness, so a modeling language like SL would not suffer from having such limitation, if it achieves it by an accurate abstraction. For example, naturally occuring (?) event sequences can cause black holes to form: the universe can give rise to patterns that eventually violate its own integrity. A modelling framework that did not have this characteristic would not be accurately represent systems of this type.

Dave Gould

Hello, George.

Happy to help in anyway possible. I developed a basic class for PhD students in Management for Walden University, read your book (ok, some of it), and will be attending the SFI summer session on sustainable cities. My background is in software engineering and systems.

George Mobus

@dk,

Yes, the issue of incompleteness or consistency is, in my opinion, irrelevant since no model ever comes close to representing the real thing. I assume this is what you mean by an "accurate abstraction," one that reflects reasonably the system behavior within the limits of accuracy and precision with which we can measure it in the first place. I have no idea about black holes - not my area of expertise.

------------------------------------
@Dave G.

My UW e-mail is gmobus@uw.edu. Drop me a line and we can take this off-line.

George

Jerry McManus

I could be wrong, but I believe the term "negative feedback" is being used incorrectly, both in the article and in the comments.

https://en.wikipedia.org/wiki/Negative_feedback

Based on the context in which the term is being used here I believe "negative consequences" would be more appropriate. My assumption being that the intent is to evoke a catastrophic collapse of society after a period of unsustainable exponential growth.

Here is an easy way that I have found to remember the difference between positive or self-reinforcing feedbacks and negative or self-limiting feedbacks: Picture a water tank with a float valve (not unlike those found on the back of most flush toilets).

If the float valve is designed to restrict the flow of water then you have a classic negative feedback, the higher the water level the slower the tank is filled, the optimal level is gradually approached, and a very stable state is eventually reached when the tank is full and the water is shut off.

Now imagine a float valve that is just the opposite, designed to increase the water flow as the water level rises. This is a classic positive feedback which leads to exponential growth of the water level. The higher the water level the more the valve opens up, before long the water overflows the tank even as the now wide open valve continues to gush water. This is a highly unstable state and the resulting catastrophic flooding of the whole house is an extremely negative consequence of a positive feedback loop.

I dunno, maybe I'm missing something, but by just about any measure the catastrophic collapse of industrial civilization and resulting die-off of the population is an extremely negative consequence of the positive feedbacks dominant in human society for at least the last couple centuries now, and in no way resembles the stable, goal-seeking behavior of negative feedbacks.

Jerry McManus

Not to belabor the point, but upon further reflection it occurred to me that the other term that would be appropriate in this context is "negative exponential" otherwise known as "exponential decay"

https://en.wikipedia.org/wiki/Exponential_decay

In other words, roughly the same dynamic as exponential growth, except in reverse. A given stock is reduced by half at each interval instead of doubling at each interval.

In the case of human society, if we reversed the exponential growth of the last two centuries which had a doubling time of approx. 40 years, then we would get an exponential decay with a half life of approx. 40 years.

Starting about now, if projections such as the ones in the Limits to Growth report are correct, then in about 40 years we will have half the population, half the industrial output per capita, half the food per capita, and half the services per capita that we do now.

See, for example, the current situation in Venezuela, or the current situation just about anywhere in Middle East / North Africa, or..., well, you get the idea.

Jerry McManus

Now that we have that sorted, lets play a fun little thought experiment. Imagine for a moment just how different our world would be right now if negative feedbacks had indeed been dominant in human society for the last couple of centuries.

For example, we know that human population has been increasing exponentially due to the well known positive feedback inherent in population dynamics: The more people there are the more babies are born. What if instead, say around 1900 or so when the population was still roughly 1 billion, the world agreed that four billion people was the highest the population could ever go and still have some expectation of all future generations living in relative comfort and security.

Well, that would obviously require a negative feedback: The more people there are the fewer babies are born. Instead of the inherently unsustainable, potentially chaotic, and almost certainly catastrophic exponential growth of the last century we would have instead seen the population gently ease up to the goal of 4 billion people on our lovely little planet and then level off. Not because people are not having any babies at all, but because people know and understand that every year there is such a thing as enough babies.

Here's another example, we know that pollution in the atmosphere is increasing exponentially due not only to our growing population, but also because of the growing standards of living that have been bought with a one time trust fund, deposited millions of years ago, of fossil sunlight stored in the Earth's crust.

What if instead, say around 1900 or so when the steam engine was well established and internal combustion was just starting to take off, the world agreed that 400ppm of carbon dioxide (or its equivalent?) in the atmosphere was the highest it could ever go and still have some expectation of all future generations living in relative comfort and security.

Well, that would obviously require a negative feedback: The more carbon emitted by human activity resident in the atmosphere then the less fossil fuel that could be burned. Instead of the inherently unsustainable, potentially chaotic, and almost certainly catastrophic climate change that we are seeing today we would have instead seen CO2 levels gently ease up to the goal of 400ppm on our magical little planet and then level off. Not because people are shivering in the cold and dark, but because people know and understand that there is such a thing as enough "blankets", and they got busy a long time ago figuring out how to live comfortably without the benefit of burning copious quantities of fossil sunlight.

It's startling to think that such a small change in the dynamics of human society, from positive feedbacks to negative feedbacks, could have such a profound effect on not only our world but also on the world of countless future generations to come. I believe Buckminster Fuller called this "Trimtab" after the steering mechanism on large ocean going vessels:

https://www.brainpickings.org/2015/08/21/buckminster-fuller-trim-tab/

And, let's not forget, Donella Meadows brilliant and life-altering essay on leverage points:

http://donellameadows.org/archives/leverage-points-places-to-intervene-in-a-system/

Oh, and BTW, someone do me a favor and let Ugo Bardi know that he also consistently misuses (I would even say abuses) the term "negative feedback". Thanks!

George Mobus

@Jerry Mc.

Thank you for your thoughtful post.

For clarification the model I am talking about is one in which initially positive feedback loops dominate pushing growth (e.g. population growth due to interactions between birth rates and population - exponential growth). But as those positive feedbacks drive the variable(s) upward, the mechanisms underlying them "feed" the rise and ultimate dominance of negative loops, e.g. deaths due to overpopulation.

So there is no real misapplication of the terms here. I've been writing about this for so long I assume readers will recognize my shorthand references. The general model of positive (growth) followed by the rise of negative (peaking and possible degrowth) is generally understood. I cover it in my Principles of Systems Science book. It is also the basis of Ugo's work.

George

Jerry McManus

Thanks George,

I won't bother splitting hairs, you say "degrowth", I say "exponential decay", obviously they both mean the same thing.

I had hoped you (and Ugo) would be a little more rigorous, especially considering the importance of the subject, but C'est la vie, as they say.

The comments to this entry are closed.