complicated v complex
complex adaptive systems (CAS)
“complex” (gardening, cultivation, farming, raising children, cooking?, nuclear reactor [also known as atomic reactor], ...)
Complicated mechanisms (clock work)
Joshua Cooper Ramo (author), The seventh sense (book), 2016
pp.136-137
p.136
“There are systems of crucial interest to humankind that have so far defied accurate simulation”, the scientist John Holland observed in a famous paper that helped establish the discipline of chaos science. Holland spent years considering these puzzling, hard-to-model systems and spotted at least one common theme Whether it was webs of finance, such as the futures exchange, or immunological networks or our own brains, highly connected systems share what Holland labeled an evolving structure ── they never stay the same. They seem to shift with an easy plasticity, in response to internal pressures or external changes. This is why so much unexpected chaos is occurring now, from government collapses to economic crises.
p.137
Connection means systems take on new forms. In many cases, they become better, stronger, more adaptively fit. It isn't simply that the unexpected appears or that there is more or less good or evil now; it's that the systems are evolving. Holland thought the world was filled with such evolution, no different from species' adjusting (or not) to a hotter climate or some fast new predator. He called the networks that produce these sorts of innovations complex adaptive systems.
“There are systems”: John Holland, “Complex Adaptive Systems”, Daedalus 121, no. 1 (Winter 1992): 17.
p.137
When Holland chose the word “complex”, he was making an important distinction. Complicated mechanisms can be designed, predicted, and controlled. Jet engines, artificial hearts, and your calculator are complicated in this sense. They may contain billions of interacting parts, but they can be laid out and repeatedly predictably made and used. They don't change. Complex systems, by contrast, can't be so precisely engineered. They are hard to fully control. Human immunology is complex in this sense. The World wide Web is complex. A rain forest is complex: It is made up of uncountable buzzing, connecting bugs and birds and trees. Order, to the extent that it exists in the Amazon basin, emerges moment by moment from countless, constant interactions. The uneven symphonic sound of l'heure bleue, that romatic stopping point at dawn when you can hear the forest waking bird by bird, is the sound of complexity engaging in a never-the-same-twice phase transition.
p.137
The word “complex” comes to us from the Latin word plexus, meaning “having parts”, which hints at the interwoven, layered nature of any object. What appears simple ── a flower, our skin, the value of a dollar bill ── is in fact a plexus, loaded with twitches and influences. In that stitching of new links, countless interactions inevitably erupt in unexpected ways: financial panics or disease epidemics or revolutions.
pp.137-138
Traffic during ruch hour is complex in this fashion ── it is the moving bits, the mishmash of cars and pedestrians and bicycles, that determine the ultimate state of the system: jammed or not.
p.138
As any system fills out with more actors and more types of connection, it becomes more complex and harder to predict. Merely complicated systems, by contrast, don't produce uncertainty in the same way; appealingly, they just run. Strapping a complicated jet engine to the wing of a passenger plane makes sense, even if it takes decades of refinement to achieve real reliability. Doing that with a complex jet engine? Not so wise.
p.138, p.139
former president of the European Central Bank Jean-Claude TRichet lamented
p.139
“As a policy maker during the crisis I found the available models of limited help. In fact, I would go further: In the face of the crisis, we felt abandoned by conventional tools.”
Joshua Cooper Ramo, The seventh sense: power, fortune, and survival in the age of network, 2016.
____________________________________
James Rickards, The death of money, 2014 [ ]
p.266
Many analysts use the words complex and complicated interchange-ably, but that is inexact. A complicated mechanism, like the clockworks can be assembled and disassembled in straightforward ways. The parts do not adapt to one another, and the clock cannot suddenly turn into a sparrow and fly away. In contrast, complex systems sometimes do morph and fly away, or slide down mountains, or ruin nations. Complex systems include moving parts, called autonomous agents, but they do more than move. The agents are diverse, connected, interactive, and adaptive. Their diversity and connectivity can be modeled to a limited extent, but interaction and adaptation quickly branch into a seeming infinity of outcomes that can be modeled in theory but not in practice. To put it another way, one can know that bad things might happen yet never know exactly why.
Clocks, watches, and motors are examples of contrained systems that are complicated but not complex. Constrast these with ubiquitous complex systems, including earthquakes, hurricanes, tornadoes--and capital markets. A single human being is a complex system. ...[..]... Risk management is possible with the right combination of complexity and another essential: humility about what is knowable.
p.269
Complexity offers a way to understand the dynamics of feedback loops through recursive functions. These have so many instantaneous iterations that explosive results may emerge from minute causes too small even to be observed.
p.269
Modern economists spend their time looking for the subatomic particle while ignoring the critical state of the system. They are looking for snowflakes and ignoring the avalanche.
(The death of money : the coming collapse of the international monetary system, James Rickards, 2014, )
____________________________________
• p.270
• Criticality is a system means that it is on the knife-edge of collapse. Not every complex system is in a critical state, as some may be stable or subcritical. One challenge for economists is that complex systems not in the critical state often behave like non-complex systems, and their stochastic properties can appear stable and predictable right up to the instant of criticality, at which point emergent properties manifest and a catastrophe unfolds, too late to stop.
• (The death of money : the coming collapse of the international monetary system, James Rickards, 2014, p.270)
____________________________________
George P. Richardson, Feedback thought in social science and systems theory, 1991 [ ]
[p.126]
TABLE 3.1 Kenneth Boulding's hierarchy of systems (abstracted from Boulding 1956, pp. 89-94)
-----------------------------------------------------------------------
Level Characteristic unit Summary description
-----------------------------------------------------------------------
(1) framework static systems
(2) clockwork simple dynamic systems
(3) thermostat control mechanisms and cybernetic systems
(4) cell open systems, or self-maintaining structures
(5) plant genetic/societal systems
(6) animal mobile, teleological systems with self-awareness
(7) human individual animal systems with self-consciousness
(8) human society social systems with self-consciousness
(9) transcendental idea ultimate, absolutes, and inescapeable knowledges
-----------------------------------------------------------------------
(Richardson, George P., Feedback thought in social science and systems theory, copyright © 1991 by the University of Pennsylvania Press)
(Feedback thought in social science and systems theory / George P. Richardson (1991), 1. social science--methodology., 2. feedback control systems., p.126 )
[p.171]
-----------------------------------------------------------------------
SYSTEMS Simple Complex Exceedingly
complex
-----------------------------------------------------------------------
Deterministic Window catch Electronic digital EMPTY
computer
--------------------------------------
Billiards Planetary system
--------------------------------------
Machine-shop Automation
lay-out
-----------------------------------------------------------------------
Probabilistic Penny tossing Stockholding The economy
---------------------------------------------------------
Jellyfish Conditioned The brain
movement reflexed
---------------------------------------------------------
Statistical Industrial THE COMPANY
quality control profitability
FIGURE 4.1: Stafford Beer's classification of systems based on degrees of complexity and uncertainty. Source: Beer (1959, p. 18).
Beer, Stafford (1959/1967). Cybernetics and Management (London: English Universities Press).
(Richardson, George P., Feedback thought in social science and systems theory, copyright © 1991 by the University of Pennsylvania Press)
(Feedback thought in social science and systems theory / George P. Richardson (1991), 1. social science--methodology., 2. feedback control systems., p.171 )
____________________________________
self-organized critical state
self-organized critical (SOC) state
self-organized criticality
earthquake
forest fire
avalanche
Per Bak, How nature works, 1996 [ ]
p.49
We had discovered a simple model for complexity in nature.
p.50
Maybe understanding is coming up with metaphoric pictures. The physics of our messy system of pendulums was far from transparent. Our intuition was poor. A couple of months after the discovery, it struck us that there was a simpler picture that could be applied to our self-organized critical dynamics. By a change of language the rotating pendulums could be describing toppling grains of sand in a pile of sand (Figure 1). Instead of counting revolutions of pendulums, we would count toppling grains at some position in the pile. Although the mathematical formulation was exactly the same for the sand model as for the pendulum model, the sand picture led to a vastly improved intuitive understanding of the phenomenon. Sandpiles are part of our everyday experience, as any child who has been playing on the beach knows. Rotating coupled pendulums are not. In a mysterious way, the physical intuition based on the sandpile metaphor leads to better understanding of the behavior of a purely mathematical model. Usually we achieve physical undestanding from mathematical analysis, not the other way around.
p.50
Consider a flat table, onto which sand is added slowly, one grain at a time. The grains might be added at random positions, or they may be added only at one point, for instance at the center of the table. The flat state represents the general equilibrium state; this state has the lowest energy, since obviously we would have to add energy to rearrange the sand to form heaps of any shape. If we had used water, the system would always return to the flat ground state as the water would simply run off the edge of the table. Because the grains tend to get stuck due to static friction, the landscape formed by the sand will not automatically revert to the ground state when we stop adding sand.
p.50
Initially, the grains of sand will stay more or less where they land. As we continue to add more sand, the pile becomes steeper, and small sand slides or avalanches occur. The grain may land on top of other grains and topple to lower level. This may in turn cause other grains to topple. The addition of a single grain of sand can cause a local disturbance, but nothing dramatic happens to the pile. In particular, events in one part of the pile do not affect sand grains in more distant parts of the pile.
p.51
There is no global communication within the pile at this stage, just many individual grains of sand.
p.51
As the slope increase, a single grain is more likely to cause other grains to topple. Eventually the slope reaches a certain value and cannot increase many further, because the amount of sand added is balanced on average by the amount of sand leaving the pile by falling off the edges. This is called a stationary state, since the average amount of sand and the average slope are constant in time. It is clear that to have this average balance between the sand added to the pile, say, in the center, and the sand leaving along the edges, there must be communication through the entire system. There will occassionally be avalanches that span the whole pile. This is the self-organized critical (SOC) state.
The additional grains of sand has transformed the system from a state in which the individual grains follow their own local dynamics to a critical state where the emergent dynamics are global.
p.51
In the stationary SOC state, there is one complex system, the sandpile, with its own emergent dynamics. The emergence of the sandpile could not have been anticipated from the properties of the individual grains.
p.51
The sandpile is an open dynamical system, since sand is added from outside. It has many degrees of freedom, or grains of sand. A grain of sand landing on the pile represents potential energy, measured as the height of the grain above the table. When the grain topples, this energy is transformed into kinetic energy. When the toppling grain comes to rest, the kinetic energy is dissipated, that is, transformed into heat in the pile. There is an energy flow through the system. The critical state can be maintained only because of energy in the form of new sand being supplied from the outside.
p.51
The critical state must be robust with respect to modifications. This is of crucial importance for the concept of self-organized criticality [SOC] to have any chance of describing the real world; in fact, this is the whole idea.
p.51
Suppose that after the same system has reached its critical stationary state we suddenly start dropping wet sand instead of dry sand. Wet sand has greater friction than dry sand. Therefore, for a while the avalanches would be smaller and local. Less material will leave the system since the small avalanches cannot reach the edge of the table. The pile becomes steeper. This, in turn, will cause the avalanches go grow, on average.
p.52
Eventually we will be back to the critical state with system-wide avalanches. The slope at this state will be higher than the original ones.
p.52
The slope at this state will be higher than the original ones. Similarly, if we dry the sand, the pile will sink to a more shallow shape by temporarily shedding larger avalanches.
p.52
If we try to prevent avalanches by putting local barriers, “snow” screens, here and there, this would have a similar effect: for a while the avalanches will be smaller, but eventually the slope will become steep enough to overcome the barriers, by forcing more sand to flow somewhere else. The physical appearance of the pile changes, but the dynamics remain critical. The pile bounces back to critical state when we try to force it away from the critical state.
p.52
The sandpile model that Kurt, Chao, and I studied is easy to define and simulate on the computer.
p.52
Readers who do not play with computers can make a mechanical representation using Lego blocks.
p.54
Nevertheless, the consequences of these rules are horrifying complicated, and can certainly not be deduced from a simple inspection of the equations, which represent the local dynamics of each of our sand grains.
pp.54-55
Second, we not particularly interested in sand. We hope that the sand dynamics that we observe are general enough that they can be applied to a much larger class of phenomena.
Peter Grassberger, a computational physicist at the University of Wupertal, Germany has come up with an amusing representation of the model. He asks us to think about a large office where bureaucrats sit at tables organized in rows (Figure 13). Every now and then a piece of paper from the outside enters the desk of a random bureaucrat. He does not deal with it until he finds too many pieces of paper on his desk. He then sends one piece of paper to each of his four neighbors. Everybody follows this rule, except those who are placed along the walls, who simply throw the paper out the window. Jumping forward a little bit, we shall see that a single piece of paper entering the office can lead to a bureaucratic catastrophe where millions of transfers of paper take place (if the office is large enough!). Each bureaucrat may perform many transactions within such an avalanche.
p.57
The straight line indicates that the avalanches follow the Gutenberg-Richter power law, just like the real earthquakes in Figure 2, although the slopes are different.
p.57
We do not have to wait millions of years to generate many earthquakes, so our statistical fluctuations are smaller than those for [real-world] earthquakes, where we must deal with the smaller number that nature has generated for us.
p.57
The power law indicates that the stationary state is critical. We conclude that the pile has self-organized into a critical state.
One can show, by analyzing the geometry of the sandpile, that the profile of the sandpile is a fractal, like Norway's coast. The avalanches have carved out fractal structures in the pile.
The power law also indicates that the distribution of avalanches follows Zipf's law.
p.57
This is just another way of representing the information from the original power law. The straight line shows that the sandpile dynamics obey Zipf's law.
p.57
Nevertheless, one might speculate that Zipf's law indicates that the world population has organized itself into a critical state, where cities are formed by avalanches of human migrations.
p.58
The power law should prevail no matter how we modify the sandpile.
p.58
We had to check that the criticality is robust with respect to modification of the model. The power law should prevail no matter how we modify the sandpile. We tried a long sequence of different versions. Instead of having the same critical height equal to 3, a version where the critical height varies from site to site was tried. Snow screens were simulated by preventing sand from falling between certain neighbor sites, selected randomly, by having the sand arranged on a triangular grid instead of square grid. We also tried adding grains of different sizes, that is, we increased Z not by unity when grains are falling but by some random number between 0 and 1. We massaged the model so that a random amount of sand topples when the site becomes unstable. We selected the sites to which the sand would topple in a random way, and not to the nearest neighbors. In all cases, the pile organized itself into a critical state with avalanches of all sizes. The criticality was unavoidable.
p.58
In all cases, the pile organized itself into a critical state with avalanches of all sizes. The criticality was unavoidable.
One might speculate that the criticality is caused by the randomness of the way that the system is driven--we add new grains at random positions. In fact, this is not important at all. We can drive the system in a deterministic way with no randomness whatever, with all information about the system at all times encoded in the initial condition: let the Zs represents a real variable instead of an integer one.
p.58
One might speculate that the criticality is caused by the randomness of the way that the system is driven ── we add new grains at random positions. In fact, this is not important at all. We can drive the system in a deterministic way with no randomness whatsoever, with all information about the system at all times encoded in the initial conditions: let the Zs represent all real variable instead of an integer one. Start with a configuration where all the Zs are subcritical, that is, less than 4. Increase all Zs at a very small rate. This corresponds to tilting the sandpile slowly. At some point, one Z will become unstable and topple according to the rule defined above, and a chain reaction is initiated. The process is continued ad infinitum; there will eventually be a balance between the rate of changing the slope and the rate of sand falling off the edges. We get the same power law distribution as before. Since the whole history of the pile in this case was contained in the initial condition, the phenomenon of SOC [Self-Organized Critical state] is essentially a deterministic phenomenon, just like the chaos studied by Feigenbaum.
p.58
The fact that the randomness of adding sand does not affect the power law indicates that the randomness is irrelevant fro the complex behavior we are observing. This fact is important to realize when studying much more complicated systems. Economics deals with the more or less random behavior of many agents, whose minds were certainly not made up at the beginning of history.
p.58
Economics deals with the more or less random behavior of many agents, whose minds were certainly not made up at the beginning of history. Nevertheless, this randomness does not preclude the system's evolving to the delicate critical state, with well-defined statistical properties.
pp.58-59
Nevertheless, this randomness does not preclude the system's evolving to the delicate critical state, with well-defined statistical properties. This is a fascinating point that is difficult to grasp. How can a system evolve to an organized state despite all the obvious randomness in the real world? How can the particular configuration be contingent on minor details, but the criticality totally robust?
p.59
In a noncritical world nothing dramatic ever happens.
p.59
Not only can he predict what will happen, but he can also understand it, to the limited extent that there is something to understand.
p.59
Once the pile has reached the stationary critical state, though, the situation is entirely different. A single grain of sand might cause an avalanche involving the entire pile. A small change in the configuration might cause what would otherwise be an insignificant event to be come catastrophe. The sand forecaster can still make short time predictions by carefully identifying the rules and monitoring his local environment. If he sees an avalanche coming, he can predict when it will hit with some degree of accuracy.
p.59
A single grain of sand might cause an avalanche involving the entire pile. A small change in the configuration might cause what would otherwise be an insignificant event to become a catastrophe. The sand forecaster can still make short time predictions by carefully identifying the rules and monitoring his local environment. If he sees an avalanche coming, he can predict when it will hit with some degree of accuracy. However, he can not predict when a large event will occur, since this is contingent on very minor details in the configuration of the entire sandpile. THe relevance of contingency in the self-organized critical state was first noted by Maya Paczuski, then a research fellow in our group, who suggested that the massive contingency in the real world could be understood as a consequence of self-organized criticality.
( Bak, P. (Per), 1947-, How nature works : the science of self-organized criticality / Per Bak., 1. critical phenomena (physics), 2. complexity (philosophy), 3. physics--philosophy., QC173.4.C74B34 1996, 003'.7--dc20, 1996, )
____________________________________
“sandpile game”
Mark Buchanan, Ubiquity, 2001 [ ]
pp.17-18
And of all these games, one stands out as a kind of archetype of simplicity, and has been central to the discovery of the underlying cause of a vast range of tumultuous events.
p.18
“sandpile game”
p.18
To understanding this “sandpile game”, a focal point for our story, imagine dropping grains of sand one by one onto a table and watching the pile grow. A grain falls accidentally here or there, and then in time the pile grows over it, freezing it in place. Afterward, the pile feels forever more the influence of that grain being just where it is and not elsewhere. In this case, clearly, history matters, since what happens now can never be washed away, but affects the entire course of the future.
p.18
“All great deeds and all great thoughts”, Albert Camus once wrote, “have ridiculous beginnings.”17 And so it was in 1987 when physicists Per Bak, Chao Tang, and Kurt Weisenfeld began playing this sandpile game in an office at Brookhaven National Laboratory, in New York State. As it turns out, this seemingly trival game lies behind the discovery of the widespread importance of the CRITICAL STATE--the discovery that can help us to make sense of upheavals.
p.18
Theoretical physicists enjoy posing seemingly trivial questions that, after a bit of thinking, turn out not to be so trivial. In this respect the sandpile game turned out to be a real winner. As grains pile up, it seems clear that a broad mountain of sand should edge slowly skyward, and yet things obviously cannot continue in this way. As the pile grows its sides become steeper, and it becomes more likely that the next falling grain will trigger an avalanche. Sand would then slide downhill to some flatter region below, making the mountain smaller, not bigger. As a result, the mountain should alternately grow and shrink, its jagged silhouette forever fluctuating.
pp.18-19
What is the typical rhythm of the growing and shrinking sandpile? Of course, they [Per Bak, Chao Tang, and Kurt Weisenfeld] didn't really care about sandpiles. In studying this silly problem, they were really chasing some insights regarding the general workings of nonequilibrium systems. The sandpile seemed like a nice, simple starting point, and with luck, they hoped, they might discover in this setting some patterns of behavior that would apply to a lot more than just sandpiles.
p.136
Things that live in CRITICAL STATES tend to show similar kinds of organization, and this organization arises not from specific details of those systems and the elements that make them up, but from the far deeper skeleton of basic geometry and logic behind these details. The critical form wells up in things regardless of what they are. When something is recognized to be in CRITICAL STATE, its essential character can be understood even by ignoring most of the details.
( Ubiquity / Mark Buchanan., 1. causality (physics), 2. pattern formation (physical science), QC6.4.C3 B83 2001, 530'.01--dc21, originally published in different form, in Great Britain, by Weidenfeld & Nicolson in 2000., 2000, 2001, )
____________________________________
Ori Brafman and Rod A. Beckstrom, The starfish and the spider, 2006 [ ]
p.47
AA is found wherever a group of members chooses to meet.
p.97
Almost every decentralized organization that has made it big was launched from a preexisting platform. Bill W., the founder of AA, drew upon the Oxford Group, an independent Christian movement started by a renegade Lutheran minister. The Oxford Group had established circles and even a 6-step program for recovery. Bill W. changed the six steps into 12, borrowed the methodology, and launched his first AA circle.
pp.46-52
1. is there a person in charge?
2. are there headquarters?
3. if you thump it on the head, will it die?
4. is there a clear division of roles?
5. if you take out a unit, is the organization harmed?
6. are knowledge and power concentrated or distributed?
7. is the organization flexible or rigid?
8. can you count the employees or participants?
9. are working groups funded by the organization, or are they self-funding?
10. do working groups communicate directly or through intermediaries?
pp.88-101
LEG 1: Circles
The only way for outsiders to join a circle, in fact, was to be taken in battle. But once brought into a circle, members were accepted as Apache--whether by birth, adoption, or capture. That's the thing about circles: once you join, you're an equal. It's then up to you to contribute to the best of your ability.
As the norms of a circle develop, and as members spend more time together, something fascinating happens: they begin to trust one another.
LEG 2: The Catalyst
The thing is, ammonia doesn't have any iron in it--it's made solely of hydrogen and nitrogen. The iron in this equation remains unchanged: it just facilitates the bonding of hydrogen and nitrogen in a certain way.
Iron is a catalyst. In chemistry, a catalyst is any element or compound that initiates a reaction without fusing into that reaction.
The catalyst is an inspirational figure who spurs others to action. Circles don't form on their own.
A catalyst develops an idea, shares it with others, and leads by example.
A catalyst is like the architect of a house: he's essential to the long-term structural integrity, but he doesn't move in.
He wasn't interested in creating an empire under his control; he was focused on sparking a movement to end slavery.
LEG 3: Ideology
Ideology is the glue that holds decentralized organization together.
The Apaches held a common belief that they belonged on the land and deserved to be self-governing. Those few Apaches who didn't hold this ideology accepted the Spanish invitation to become farmers and integrate into a centralized system. But those who stayed with the tribe held firmly to the notion of independence. Anyone who interfered with that ideology--whether a Spaniard, a Mexican, or an American--became the enemy. The Apaches held to their ideology so strongly that they were willing to fight and sacrifice themselves for their cause.
LEG 4: The Preexisting Network
The Quakers
the Quakers gave the movement a platform
Third, and most important, centralized organizations aren't set up to launch decentralized movements.
... slowly gained their trust and friendship.
LEG 5: The Champion
A champion is relentless in promoting a new idea.
Leor Jacobi
Just ask the folks at the Berkeley post office in California--they're still talking about Leor Jacobi.
Something about the way Leor spoke--his excitement or his charm--made everyone feel comfortable with him and interested in what he had to say.
Champions are inherently hyperactive.
(Brafman, Ori, The starfish and the spider : the unstoppable power of leaderless organization / Ori Brafman and Rod A. Beckstrom., 1. decentralization in management., 2. organization behavior., 3. success in business., 2006, )
____________________________________
No comments:
Post a Comment