What do you think?
Rate this book
Understanding Complexity
Length: 6hrs 4min
Recent years have seen the introduction of concepts from the new and exciting field of complexity science that have captivated the attention of economists, sociologists, engineers, businesspeople, and many others. These include tipping points, the wisdom of crowds, six degrees of separation (or Kevin Bacon), and emergence.
Interest in these intriguing concepts is widespread because of the utility of this field. Complexity science can shed light on why businesses or economies succeed and fail, how epidemics spread and can be stopped, and what causes ecological systems to rebalance themselves after a disaster.
In fact, complexity science is a discipline that may well hold the key to unlocking the secrets of some of the most important forces on Earth. But it's also a science that remains largely unknown, even among well-educated people.
Now you can discover and grasp the fundamentals and applications of this amazing field with Understanding Complexity. Professor Scott E. Page of the University of Michigan - one of the field's most highly regarded teachers, researchers, and real-world practitioners - introduces you to this vibrant and still evolving discipline. In 12 lucid lectures, you learn how complexity science helps us understand the nature and behavior of systems formed of financial markets, corporations, native cultures, governments, and more.
PLEASE NOTE: When you purchase this title, the accompanying PDF will be available in your Audible Library along with the audio.
©2009 The Great Courses (P)2009 The Teaching Company, LLC
Recent years have seen the introduction of concepts from the new and exciting field of complexity science that have captivated the attention of economists, sociologists, engineers, businesspeople, and many others. These include tipping points, the wisdom of crowds, six degrees of separation (or Kevin Bacon), and emergence.
Interest in these intriguing concepts is widespread because of the utility of this field. Complexity science can shed light on why businesses or economies succeed and fail, how epidemics spread and can be stopped, and what causes ecological systems to rebalance themselves after a disaster.
In fact, complexity science is a discipline that may well hold the key to unlocking the secrets of some of the most important forces on Earth. But it's also a science that remains largely unknown, even among well-educated people.
Now you can discover and grasp the fundamentals and applications of this amazing field with Understanding Complexity. Professor Scott E. Page of the University of Michigan - one of the field's most highly regarded teachers, researchers, and real-world practitioners - introduces you to this vibrant and still evolving discipline. In 12 lucid lectures, you learn how complexity science helps us understand the nature and behavior of systems formed of financial markets, corporations, native cultures, governments, and more.
PLEASE NOTE: When you purchase this title, the accompanying PDF will be available in your Audible Library along with the audio.
©2009 The Great Courses (P)2009 The Teaching Company, LLC
7 pages, Audible Audio
First published January 1, 2009
42 people are currently reading
1498 people want to read
Ratings & Reviews
Friends & Following
Create a free account to discover what your friends think of this book!
Community Reviews
Displaying 1 - 30 of 107 reviews
August 9, 2019
An outstanding series of lectures. One of the best courses I've ever taken, both in delivery and contents. I might be biased, since I was already extremely interested in complexity. However, it is hard not to be impressed by the structure of the lectures and the enthusiasm of the speaker. Highly recommended.
October 21, 2019
I picked this one up because Scott Page is an old college friend, and hearing his voice – not so much as a reader as a lecturer – was part of the deep pleasure of listening to it. I remember him back in his student government days with the same mix of seriousness and good humor – putting energy into the work but never taking himself too seriously.
So, biased as I am, I declare his presence, his voice, worth the price of admission.
But there’s much more here as well. I guess I could have provided a workable definition of what it means for something to be complex, but I’d never have been able to weave a basically simple concept into the playful depths that Scott manages.
The fundamental observation here is that “complexity” represents a space between easily mappable scenarios and what seems pure randomness. We can measure an economic exchange that involves a couple producers and a couple consumers. We have no hope of measuring something as unpredictable as quantum motion. In between lies something like the macroeconomic conditions we know, an area we cannot accurately predict or control but that we do have the power to influence.
Understanding that in-between, “understanding complexity,” is a dramatic new frontier in data science. It’s the child of game theory and chaos theory, suggesting that simple concepts, twined together, can produce a complex view of what we mean by complexity itself. It feels like a wonderful logic game, and it also feels like it might be the key to making our world substantially better.
I’m not doing justice either to the substance or color of Scott’s argument, but it’s stimulating throughout. He has wonderful metaphors like “dancing landscapes” and “Mt. Fujis,” and he has a knack for setting up the concepts early that he will need later.
My favorite part here, I think, is the way he demonstrates the power of agent-based model simulations. He demonstrates throughout this that, when we account for “bottom-up” phenomena of organized systems – which is any system where potentially countless individual actors make individually based determinations that produce a potentially predictable reaction in an interdependent whole – bizarre and wonderful things can happen. He gives the example of computer simulations in which cells light up as black or white when they meet certain conditions (such as whether their neighbors are black or white) and then produce seemingly top-down results – such as when a set of black/white binary cells produce what looks like a stick-figure creature taking steps forward.
I’m rushing to get down ideas that Scott made clear in sustained fashion but that came to me in spasms of understanding. I’m not deleting this one because I am tempted already to listen to it again.
In the meantime, I feel smarter for having listened to it.
So, biased as I am, I declare his presence, his voice, worth the price of admission.
But there’s much more here as well. I guess I could have provided a workable definition of what it means for something to be complex, but I’d never have been able to weave a basically simple concept into the playful depths that Scott manages.
The fundamental observation here is that “complexity” represents a space between easily mappable scenarios and what seems pure randomness. We can measure an economic exchange that involves a couple producers and a couple consumers. We have no hope of measuring something as unpredictable as quantum motion. In between lies something like the macroeconomic conditions we know, an area we cannot accurately predict or control but that we do have the power to influence.
Understanding that in-between, “understanding complexity,” is a dramatic new frontier in data science. It’s the child of game theory and chaos theory, suggesting that simple concepts, twined together, can produce a complex view of what we mean by complexity itself. It feels like a wonderful logic game, and it also feels like it might be the key to making our world substantially better.
I’m not doing justice either to the substance or color of Scott’s argument, but it’s stimulating throughout. He has wonderful metaphors like “dancing landscapes” and “Mt. Fujis,” and he has a knack for setting up the concepts early that he will need later.
My favorite part here, I think, is the way he demonstrates the power of agent-based model simulations. He demonstrates throughout this that, when we account for “bottom-up” phenomena of organized systems – which is any system where potentially countless individual actors make individually based determinations that produce a potentially predictable reaction in an interdependent whole – bizarre and wonderful things can happen. He gives the example of computer simulations in which cells light up as black or white when they meet certain conditions (such as whether their neighbors are black or white) and then produce seemingly top-down results – such as when a set of black/white binary cells produce what looks like a stick-figure creature taking steps forward.
I’m rushing to get down ideas that Scott made clear in sustained fashion but that came to me in spasms of understanding. I’m not deleting this one because I am tempted already to listen to it again.
In the meantime, I feel smarter for having listened to it.
September 30, 2019
אין לי הרבה מה לאמר על הספר כי חלקים שלו לא הבנתי עד הסוף. גם כשהבנתי בזמן ההקשבה, אני לא מסוגלת לשחזר את דברים במרחק הזמן, כי ההבנה שלי היתה נקודתית ולא משהו קוהרנטי.
אני בטוחה שזו אני, צריך רקע מסויים במודלים מתמטים וסטטיסטיים, דבר שאין לי. הוא דן בנושאים מאוד מעניינים כמו אחידות, שונות, מגוון, אנטרופיה, מורכבות ועוד. אבל שוב, אני לא יודעת להסביר חלק מהדברים שהבנתי באופן אינטואיטיבי כשהוא הסביר אותם.
הבעיה היא שזה אודיו של קורס ואין ספר שניתן לחזור ולעקוב אחרי ההסברים. בנוסף חלק מהפרקים/ הרצאות לא עיניינו אותי.
אני בטוחה שזו אני, צריך רקע מסויים במודלים מתמטים וסטטיסטיים, דבר שאין לי. הוא דן בנושאים מאוד מעניינים כמו אחידות, שונות, מגוון, אנטרופיה, מורכבות ועוד. אבל שוב, אני לא יודעת להסביר חלק מהדברים שהבנתי באופן אינטואיטיבי כשהוא הסביר אותם.
הבעיה היא שזה אודיו של קורס ואין ספר שניתן לחזור ולעקוב אחרי ההסברים. בנוסף חלק מהפרקים/ הרצאות לא עיניינו אותי.
August 8, 2020
This is organized as a set of lectures and is a great introduction to complexity science. This is the first time I listened to a lecture set and I like it as an additional learning tool. I was looking to put structure to my thinking on complexity as a concept and this set of lectures helped me immensely.
November 5, 2019
Not to mention interesting topic, this is probably one of the most amazing learning experiences I ever had.
October 6, 2023
As a lover of the intangible that exists, I enjoyed this book. The author shows his skills as a teacher in this masterclass, in my case listened to in audiobook format.
The phenomenon of complex systems is fascinating. They exist, they are long-lasting, largely unpredictable (at least currently)... For example, the global climate, social systems, the human brain, all are complex systems. Perhaps one day we will be able to advance further in our knowledge.
Like any popular science book that meets my expectations of knowing a certain topic, I must give it five stars.
The phenomenon of complex systems is fascinating. They exist, they are long-lasting, largely unpredictable (at least currently)... For example, the global climate, social systems, the human brain, all are complex systems. Perhaps one day we will be able to advance further in our knowledge.
Like any popular science book that meets my expectations of knowing a certain topic, I must give it five stars.
September 23, 2019
This book was FASCINATING!! Highly recommended, but with one big caveat.
The caveat is that this book isn't actually about complexity but rather about complex systems. That might seem like a minor distinction but on the heels of Big History: The Big Bang, Life On Earth, And The Rise Of Humanity I've been wrestling ever since with his contention that the arc of existence is toward ever greater complexity. A comment from someone in the past year or two noting that computer science is really the study of complexity left me more curious about the topic as a whole, and thus attracted to this title. As such it was somewhat disappointing to discover that the topic as a whole wasn't really addressed here but rather was narrowed to focus on complex systems and consider that the entirety of the topic. That might indeed be a part but the arc of history in Big History or the ever-growing hierarchies of complexity and the effort to understand/manage them that CS deals with didn't feel like they were included here, and that's a big difference. I'm still looking for the book(s) for that; if anyone knows of them please comment.
That said, a book about complex systems was also interesting to me and the execution of it was, simply, superb. The topic was fairly wide ranging in the areas it touched and did a really excellent job of discussing the details at a level that was understandable without overly simplifying (e.g. the discussion of simulated annealing was really masterfully executed; technical enough to speak to the mathematical/physical details involved while still able to be comprehensible to a non-technical reader; really impressive). And the relevance/importance of the topic were well-explored and justified. Combined with how wide ranging this was and the excellent selection of examples across fields and the book felt both deeply informative and a real pleasure to listen to. I may repeat the entire thing in the near future just for the pleasure of it.
The end result being that I'd rank this among the best of the Great Courses books I've read and highly recommend it to anyone who has even a passing interest, and that despite it not speaking to the subject I came to it to learn about. Amazing.
The caveat is that this book isn't actually about complexity but rather about complex systems. That might seem like a minor distinction but on the heels of Big History: The Big Bang, Life On Earth, And The Rise Of Humanity I've been wrestling ever since with his contention that the arc of existence is toward ever greater complexity. A comment from someone in the past year or two noting that computer science is really the study of complexity left me more curious about the topic as a whole, and thus attracted to this title. As such it was somewhat disappointing to discover that the topic as a whole wasn't really addressed here but rather was narrowed to focus on complex systems and consider that the entirety of the topic. That might indeed be a part but the arc of history in Big History or the ever-growing hierarchies of complexity and the effort to understand/manage them that CS deals with didn't feel like they were included here, and that's a big difference. I'm still looking for the book(s) for that; if anyone knows of them please comment.
That said, a book about complex systems was also interesting to me and the execution of it was, simply, superb. The topic was fairly wide ranging in the areas it touched and did a really excellent job of discussing the details at a level that was understandable without overly simplifying (e.g. the discussion of simulated annealing was really masterfully executed; technical enough to speak to the mathematical/physical details involved while still able to be comprehensible to a non-technical reader; really impressive). And the relevance/importance of the topic were well-explored and justified. Combined with how wide ranging this was and the excellent selection of examples across fields and the book felt both deeply informative and a real pleasure to listen to. I may repeat the entire thing in the near future just for the pleasure of it.
The end result being that I'd rank this among the best of the Great Courses books I've read and highly recommend it to anyone who has even a passing interest, and that despite it not speaking to the subject I came to it to learn about. Amazing.
November 9, 2021
If you asked me to explain this to you I would probably just cry but it was really fun to learn about
December 18, 2019
So. This was complicated. I'm not sure if I feel smarter or stupider now. Perhaps a bit of both, so I think that means these lectures were effective.
December 10, 2025
For me probaby 3.5*s since I was actually looking for information focused more on complexity wrt math and physics, and covering topics like fractals, cellular automata, or computational complexity without venturing into human systems.
The field of complexity does include interdisciplinary applications of concepts like emergence, networks, and adaptive systems. And this class, while it does explore dynamical systems and chaos theory, extends to include the social sciences and economics.
Given that, the author and class was very good. I listened to the audiobook, often finding listening at .9x speed preferable.
The field of complexity does include interdisciplinary applications of concepts like emergence, networks, and adaptive systems. And this class, while it does explore dynamical systems and chaos theory, extends to include the social sciences and economics.
Given that, the author and class was very good. I listened to the audiobook, often finding listening at .9x speed preferable.
May 2, 2022
Scott Page brings a well organized passion to this introduction to complex systems. He begins by describing the difference between simple, complicated, and complex systems. Throughout he will use the analogy of Mount Fiji, a mountain range, and a fluctuating mountain range. In a simple system, there is a simple solution and it will take you right to the top. In a complicated system, the solution you are pursuing might take you to a localized peak, but it will not necessarily get you to the highest possible peak. The problem is a challenging but stagnant one. A complex system is like a mountain range where the valleys today might be peaks tomorrow and vice versa. The solution you are pursuing today might end up being the absolute worst solution tomorrow because outside influences are constantly changing the "landscape" of the problem.
This is how Page works. He will create such an illustration and then demonstrate how it applies in real-world examples from economics, biology, sociology, etc. This is a great way to give someone a cursory understanding of complex (pardon the pun) issues and in this series, Page proves himself the master of it. Beyond that, both his expertise and passion for the subject shine through from cover to cover. Although most won't be much interested in the issue, for those that are, this is a great intro or refresher.
This is how Page works. He will create such an illustration and then demonstrate how it applies in real-world examples from economics, biology, sociology, etc. This is a great way to give someone a cursory understanding of complex (pardon the pun) issues and in this series, Page proves himself the master of it. Beyond that, both his expertise and passion for the subject shine through from cover to cover. Although most won't be much interested in the issue, for those that are, this is a great intro or refresher.
October 30, 2025
Servicable overview on complexity theory, albeit very basic in substance and style; in particular the discussion of the phenomenon of the 'adjacent possible' was informative and inspiring; for more on the subject be sure to check out the magisterial Scale: The Universal Laws of Growth, Innovation, Sustainability, and the Pace of Life in Organisms, Cities, Economies, and Companies.
February 6, 2022
One of my favourite Great Courses. There's lots in here that's both practical and explanatory. I think tipping points are crucial to understand in many different areas of life, and diversity isn't just about ethics, it's logistically best practice in a complex system.
2022-02-06
Second time through, still good. I realised that I first listened to this in December 2019 and he talks a bit about global pandemics. 😬
2022-02-06
Second time through, still good. I realised that I first listened to this in December 2019 and he talks a bit about global pandemics. 😬
July 1, 2019
This is a nice, brief introduction to the subject. Explains many concepts like self organised criticality, agent based modeling, phase transitions, tipping points, emergence etc.. in complex systems. Highly recommend
January 19, 2020
The Great Courses really probably ought to be re-titled, "Pretty Decent 101 Courses."
August 20, 2020
Really interesting introduction into complex systems and the way of thinking.
Mapping Mount Fuji into real life, agent-based models, crashes, the importance of slack in certain systems, optimization, and a ton of other useful information.
The author/narrator is great as well. I've already looked up his books for further reading.
Mapping Mount Fuji into real life, agent-based models, crashes, the importance of slack in certain systems, optimization, and a ton of other useful information.
The author/narrator is great as well. I've already looked up his books for further reading.
March 22, 2020
Very good primer!
September 1, 2022
Good format to getting to know the subject.
January 8, 2020
Watch a detailed review along with my favorite ideas and takeaways at:
https://youtu.be/GLNGK-RuvL8
https://youtu.be/GLNGK-RuvL8
February 17, 2022
I took this course on Audible, and I loved it, it should be something that is mandatory for High School, and most definitely college. The understanding of complex systems is something that, in my opinion, is increasingly necessary for society.
June 26, 2022
What is the book about?
Understanding Complexity introduces to the framework of complex adaptive systems and its core terms and concepts. Which properties makes complex systems complex? Which properties, or rather, which degrees of property manifestation separate them from other types of systems? How can the landscape analogy be helpful to understand complex systems? — and other questions are answered by Scott Page on a low-resolution level not to overwhelm the beginner.
Can I recommend it? Whom is this book for?
Yes, indeed I can, given that you are unfamiliar with complex adaptive systems.
But first, let me tell you what I did not like.
- Sometimes Scott Page overdid on the analogies by not even finishing the first before he goes on to the next. He describes one analogical case, asks how can we understand it, and then uses another analogy. In the end it happened it did not became clear how some of the analogies can be explained in terms of the concept under discussion.
- Some definitions were given by using the term to be explained: e.g. connectedness
However, what I particularly liked was that
- many real-world analogies were used which enhanced understanding a lot and compensated for the abstraction of complex systems.
- he repeated himself. Some terms were explained repeatedly not only in compressed space (directly in sequence) but also in expanded space (across chapters) to reinforce terms and concepts.
- a very sensible chapter division was performed.
That some questions about details were skipped is forgivable considering it is a book for absolute beginners. It was interesting, very comprehensibly written, and exceeded my current level of knowledge so I could learn a lot.
I was even surprised: I expected complexity theory to be less advanced, but it is well known that complex systems were determined to possess a set of clearly defined properties (diversity, adaptiveness, interdependence, connectedness).
What did I learn?
Some facts:
- Tailorism: Named after Frederick Taylor who optimised work processes. Also: Scientific management.
- 1/4th of nodes with n+1 degree if n is the degree
- Power law distributions have lots of nodes with very low degrees, but few nodes with very high degrees (e.g. the world wide web)
- Power law networks are very robust because random node removal is more likely to remove nodes with low degrees (internal)
- Interactions can almost always be described in terms of positive and negative feedbacks
- Diversity promotes phase transitions: Diverse action readiness among the agents, that is diverse thresholds to act, establish a cascade of events
- Lags can cause a system to run in circles instead of stabilising
- Many real-world random events are not truly random: They are the outcomes of complex adaptive systems.
- Complex systems produce novelties.
- System dynamics ignores place and diversity
- The canonical standard-decision making model does not work to figure out what to do in a complex system. It does not take into account the behavior of other interested agents, it translates complexity into uncertainty, it is all exploitation, it focuses on a single outcome and not on system properties. ⇒ takes no account of what the system might be like as a result of one’s action
- Game theory either considers two players or all players, that is low connectedness or high connectedness, because it is easier to predict
- systems are sensitive to its initial conditions + rules
Key terms and concepts
Agent-based models: They are mathematical models that allow us to simulate the trajectories of a system with pre-defined factors.
- For an agent to act, some variable must be above an threshold.
The Exploration-Exploitation distinction: The trade-off between seraching for better solutions and taking advantage of what is known.
- Evolution adjusts the temperature and in turn the flexibility of the agents by changing variation in the species. (Examples of exploration: recombination, mutation; Examples of exploitation: selection)
- in rugged landscapes it is effective to first explore a lot, reduce exploration over time but increase exploitation until the peak is reached
- in dancing landscapes it is necessary to keep exploring to some degree, because one cannot keep exploiting; what was once a good solution has ceased to be one
- If exploration is low because the agents exploit, it pays off to explore better alternative solutions; if exploration is high because hardly any agent exploits, it pays off to exploit oneself
- In other words, exploitation of known solutions makes others adapt by exploring better or alternative solutions. In turn exploration makes others adapt by taking the opportunity to exploit known solutions. This keeps the landscape dancing.
Nonstationary Processes: Processes in which the probability of events changes over time. Stationary processes do not change over time
- Example: A frictionless pendulum will maintain its frequency and amplitude, but if a force is applied like friction with air frequency and amplitude will not remain constant.
Power-Law Network: A network in which the distribution of links fits a model of a type of long-tailed distribution.
Simulated Annealing: A search algorithm in which the probability of making an error decreases over time.
Coupled oscillation: Two agents changing their behavior in regular patterns in dependence on one another. While the pace of the red pendulum increases, the pace of the green pendulum decreases.
Breaking symmetry: Many individuals starting out to perform identical tasks, but then performing idiosyncratic ones.
- Central to emergence
Criticality: Property of a complex system that is prone to produce large events in response to small events by cascading.
- Tipping Point/Critical threshold: The configuration in a complex system in which a sequence of small events (cascade) can push the system into a new macro state.
- Tipping goes only in one direction: Once stability has been reached, it is hard to go back.
Phase Transition: The process in which the system moves from one macro-state to to another.
- Diversity and positive feedbacks promote phase transitions.
Feedback: interactions between instances of the same action
1. Positive feedbacks lead to tipping points/major events (by cascading), if the agents are diverse which makes their critical thresholds diverse
-Example: If everyone had the same threshold of 10, all would leave at the same time, but if one person has a threshold of 0 and another of 1 and 2, etc., then a cascade is initiated causing mass exodus despite the average threshold being, say 50 if the number of people in the room is 100. The threshold does not have to be n+1 assuming all people maintain independent variables being able to cause leaving.
- Example: segregration, despite the agents being tolerant at the micro-level.
- What inititates the cascade is a threshold at the extreme (tail of the distribution).
2. Negative feedbacks lead to stability (with changes but without major events), if the agents are diverse or their thresholds are diverse
- Examples: Lakes and bees are diverse and are able to stabilise the system, but once a certain threshold is reached, whether that is a certain amount of nitrogen in the lake or a certain temperature, the agents cannot compensate anymore for the impact caused by the external event. The lake becomes eutrophic and the bees die.
- The negative feed here is that the more nitrogen is taken up or the more the bees flap their wings, the lower the amount of nitrogen in the lake or the lower the temperature.
Externality: feedback across differnt actions (as opposed to feedback)
Long-Tailed Distribution: A distribution in which most event sizes are small but some are very large. The power law purports that one quantity varies as a power of the other.
What types of systems exist?
Complex systems are only interesting if it’s 4 main features meet a certain configuration which is marked by neither being too simple, nor too complex (chaotic). Then they become a perpetual source of novelty.
- Class 1: Stable, single point equilibria (e.g. ball resting at the bottom of a bowl, farmer’s market)
- Class 2: Periodic orbits; stable predictable patterns (e.g. stoplights, the earth rotating around the sun, prey-predator population sizes)
- Class 3: Chaotic; extremely sensitive to initial conditions (e.g. butterflies causing hurricanes)
- Class 4: Complex; somewhat stable with regular structures, but longer patterns with high information content and still difficult to predict.
What are the properties of complex systems?
1. interdependence:
- low → some change
- high → incomprehensible mangle
- moderate → complexity
- Example: deciding how to dress
2. connectedness
- low or high → quickly achieved equilibrium, because in the first case there is little to adapt to and in the second large convergence on what is most common happens
- moderate → complexity: it takes a long time to stabilise
- Example: The Greeting Game, Paper-Scissors-Rocks, sensitive, toxic, resistant E. Coli
3. diversity
- no diversity → nothing happens
- low to moderate → complexity
- high + much interdependence → incomprehensible mangle or collapse
- high + low interdependence → complexity
- Example: cooking a soup, reactions between chemical elements
4. adaptiveness
- no adaptiveness → complexity possible (Game of Life), but unlikely
- moderate → complexity: following the rules well, but not optimally
- high (following the rules optimally) → equilibrium
- Example: Checkers
What are complex systems?
They are characterised by 4 main features which are moderately pronounced:
- connected: The entities involved are connected with each other.
- interdependent: The connected entities interact with each other and influence each other, locally or globally.
- diverse: They possess entities of different types. (As opposed to variable: a difference in the value of an attribute.)
- adaptive/learning/intelligent: Their entities adapt their behavior to changes of other components in the system.
What is the difference between connectedness and interdependence?
Interdependence requires connectedness, but exerts different effect sizes depending on the degree of connectedness:
- High connectedness + low interdependence = no big effect
- High connectednesss + high interdependence = big effect
- Low connectedness + low interdependence = some effect
- Low connectedness + High interdepedence = moderate effect
What other features do complex systems exhibit?
- Selection: A process through which less fit or lower-performing entities are removed from the population.
- Robustness: The ability of a complex system to maintain functionality given a disturbance.
- Emergence: A higher-level phenomenon that arises from the micro-level interactions. (Weak: explicable; Strong: unexplicable yet, e.g. consciousness)
- Self-Organisation: A form of emergence in which the entities create a pattern or structure from the bottom-up.
- Self-Organised Criticality: A phenomenon in which interaction agents self-organise into states that can produce large events.
In contrast, what are complicated systems?
- connected
- interdependent
- diverse
- but not adaptive, not selective, not producing large events
What is it about the landscape analogy?
The landscape represents a problem or the fixed state of a complex system
- Encoding: The kind of landscape a problem or system produces, depends on the number of variables encoded. The length (and width) encode the variables. They describe the type of solution; the attribute(s) of the solution (e.g. shovel size)
- Height/Elevation represents the value of a potential solution. Local and global peaks represent (sub)optimal solutions.
- If the landscape becomes “bigger”, the number of possible solutions and non-solutions increases, and thus potentially the difficulty to solve the problem
What types of landscapes exist?
- simple landscapes: Neither interactions between actions of a single entity, nor interactions between multiple entities. They possess one global peak. (Example: optimisation of one design feature of one coal shovel.)
- rugged landscapes: Characterised by many possible combinations, and interactions within the actions of a single entity. They possess many local peaks with one (or multiple) global peaks. (Example: optimisation of all design features of one coal shovel; finding the optimal shovel from many good shovels.)
- dancing landscapes: Characterised by *externalities*, that is interactions between mulitple agents. They possess constantly changing local and global peaks. What is a solution today, might not be a solution tomorrow. They lie in the "interesting in-between": between a statistic regularity of high degrees to which the properties are pronounced and stasis of low degrees of property pronunciation. (Example: society)
How do we control complex systems?
- We build in some slack to increase robustness: Since the more a system is optimised, the more critical it becomes.
- We encourage diversity to balance exploration and exploitation, and prevent error.
- We perform safety measures for large events.
- We carefully define goals and incentives to control selection mechanisms:
- Connectedness: We create synergistic links and cut those that limits responsiveness.
What are the causes of diversity?
1. Diversity itself: the more you start with, the more combinations of the attributes of types are possible (positive feedback loop)
2. Weak selective pressure: The less types are removed from the population, the more types survive and reproduce.
3. Dancing landscapes and different landscapes:: An ever changing landscape produces multiple different landscapes that allow different types to solve problems.
What are the differences between evolutionary systems and creative systems?
1. Leap size: Evolution makes no leaps (all structures evolve slowly), while creative systems can jump forth.
2. Interim viability: Evolutionary outputs must be viable each step along the way to survive, while creative systems can produce outputs whose malfunction can be improved.
3. Representation: Evolution is fixed on genes, while creative systems can switch new ways of encoding.
Understanding Complexity introduces to the framework of complex adaptive systems and its core terms and concepts. Which properties makes complex systems complex? Which properties, or rather, which degrees of property manifestation separate them from other types of systems? How can the landscape analogy be helpful to understand complex systems? — and other questions are answered by Scott Page on a low-resolution level not to overwhelm the beginner.
Can I recommend it? Whom is this book for?
Yes, indeed I can, given that you are unfamiliar with complex adaptive systems.
But first, let me tell you what I did not like.
- Sometimes Scott Page overdid on the analogies by not even finishing the first before he goes on to the next. He describes one analogical case, asks how can we understand it, and then uses another analogy. In the end it happened it did not became clear how some of the analogies can be explained in terms of the concept under discussion.
- Some definitions were given by using the term to be explained: e.g. connectedness
However, what I particularly liked was that
- many real-world analogies were used which enhanced understanding a lot and compensated for the abstraction of complex systems.
- he repeated himself. Some terms were explained repeatedly not only in compressed space (directly in sequence) but also in expanded space (across chapters) to reinforce terms and concepts.
- a very sensible chapter division was performed.
That some questions about details were skipped is forgivable considering it is a book for absolute beginners. It was interesting, very comprehensibly written, and exceeded my current level of knowledge so I could learn a lot.
I was even surprised: I expected complexity theory to be less advanced, but it is well known that complex systems were determined to possess a set of clearly defined properties (diversity, adaptiveness, interdependence, connectedness).
What did I learn?
Some facts:
- Tailorism: Named after Frederick Taylor who optimised work processes. Also: Scientific management.
- 1/4th of nodes with n+1 degree if n is the degree
- Power law distributions have lots of nodes with very low degrees, but few nodes with very high degrees (e.g. the world wide web)
- Power law networks are very robust because random node removal is more likely to remove nodes with low degrees (internal)
- Interactions can almost always be described in terms of positive and negative feedbacks
- Diversity promotes phase transitions: Diverse action readiness among the agents, that is diverse thresholds to act, establish a cascade of events
- Lags can cause a system to run in circles instead of stabilising
- Many real-world random events are not truly random: They are the outcomes of complex adaptive systems.
- Complex systems produce novelties.
- System dynamics ignores place and diversity
- The canonical standard-decision making model does not work to figure out what to do in a complex system. It does not take into account the behavior of other interested agents, it translates complexity into uncertainty, it is all exploitation, it focuses on a single outcome and not on system properties. ⇒ takes no account of what the system might be like as a result of one’s action
- Game theory either considers two players or all players, that is low connectedness or high connectedness, because it is easier to predict
- systems are sensitive to its initial conditions + rules
Key terms and concepts
Agent-based models: They are mathematical models that allow us to simulate the trajectories of a system with pre-defined factors.
- For an agent to act, some variable must be above an threshold.
The Exploration-Exploitation distinction: The trade-off between seraching for better solutions and taking advantage of what is known.
- Evolution adjusts the temperature and in turn the flexibility of the agents by changing variation in the species. (Examples of exploration: recombination, mutation; Examples of exploitation: selection)
- in rugged landscapes it is effective to first explore a lot, reduce exploration over time but increase exploitation until the peak is reached
- in dancing landscapes it is necessary to keep exploring to some degree, because one cannot keep exploiting; what was once a good solution has ceased to be one
- If exploration is low because the agents exploit, it pays off to explore better alternative solutions; if exploration is high because hardly any agent exploits, it pays off to exploit oneself
- In other words, exploitation of known solutions makes others adapt by exploring better or alternative solutions. In turn exploration makes others adapt by taking the opportunity to exploit known solutions. This keeps the landscape dancing.
Nonstationary Processes: Processes in which the probability of events changes over time. Stationary processes do not change over time
- Example: A frictionless pendulum will maintain its frequency and amplitude, but if a force is applied like friction with air frequency and amplitude will not remain constant.
Power-Law Network: A network in which the distribution of links fits a model of a type of long-tailed distribution.
Simulated Annealing: A search algorithm in which the probability of making an error decreases over time.
Coupled oscillation: Two agents changing their behavior in regular patterns in dependence on one another. While the pace of the red pendulum increases, the pace of the green pendulum decreases.
Breaking symmetry: Many individuals starting out to perform identical tasks, but then performing idiosyncratic ones.
- Central to emergence
Criticality: Property of a complex system that is prone to produce large events in response to small events by cascading.
- Tipping Point/Critical threshold: The configuration in a complex system in which a sequence of small events (cascade) can push the system into a new macro state.
- Tipping goes only in one direction: Once stability has been reached, it is hard to go back.
Phase Transition: The process in which the system moves from one macro-state to to another.
- Diversity and positive feedbacks promote phase transitions.
Feedback: interactions between instances of the same action
1. Positive feedbacks lead to tipping points/major events (by cascading), if the agents are diverse which makes their critical thresholds diverse
-Example: If everyone had the same threshold of 10, all would leave at the same time, but if one person has a threshold of 0 and another of 1 and 2, etc., then a cascade is initiated causing mass exodus despite the average threshold being, say 50 if the number of people in the room is 100. The threshold does not have to be n+1 assuming all people maintain independent variables being able to cause leaving.
- Example: segregration, despite the agents being tolerant at the micro-level.
- What inititates the cascade is a threshold at the extreme (tail of the distribution).
2. Negative feedbacks lead to stability (with changes but without major events), if the agents are diverse or their thresholds are diverse
- Examples: Lakes and bees are diverse and are able to stabilise the system, but once a certain threshold is reached, whether that is a certain amount of nitrogen in the lake or a certain temperature, the agents cannot compensate anymore for the impact caused by the external event. The lake becomes eutrophic and the bees die.
- The negative feed here is that the more nitrogen is taken up or the more the bees flap their wings, the lower the amount of nitrogen in the lake or the lower the temperature.
Externality: feedback across differnt actions (as opposed to feedback)
Long-Tailed Distribution: A distribution in which most event sizes are small but some are very large. The power law purports that one quantity varies as a power of the other.
What types of systems exist?
Complex systems are only interesting if it’s 4 main features meet a certain configuration which is marked by neither being too simple, nor too complex (chaotic). Then they become a perpetual source of novelty.
- Class 1: Stable, single point equilibria (e.g. ball resting at the bottom of a bowl, farmer’s market)
- Class 2: Periodic orbits; stable predictable patterns (e.g. stoplights, the earth rotating around the sun, prey-predator population sizes)
- Class 3: Chaotic; extremely sensitive to initial conditions (e.g. butterflies causing hurricanes)
- Class 4: Complex; somewhat stable with regular structures, but longer patterns with high information content and still difficult to predict.
What are the properties of complex systems?
1. interdependence:
- low → some change
- high → incomprehensible mangle
- moderate → complexity
- Example: deciding how to dress
2. connectedness
- low or high → quickly achieved equilibrium, because in the first case there is little to adapt to and in the second large convergence on what is most common happens
- moderate → complexity: it takes a long time to stabilise
- Example: The Greeting Game, Paper-Scissors-Rocks, sensitive, toxic, resistant E. Coli
3. diversity
- no diversity → nothing happens
- low to moderate → complexity
- high + much interdependence → incomprehensible mangle or collapse
- high + low interdependence → complexity
- Example: cooking a soup, reactions between chemical elements
4. adaptiveness
- no adaptiveness → complexity possible (Game of Life), but unlikely
- moderate → complexity: following the rules well, but not optimally
- high (following the rules optimally) → equilibrium
- Example: Checkers
What are complex systems?
They are characterised by 4 main features which are moderately pronounced:
- connected: The entities involved are connected with each other.
- interdependent: The connected entities interact with each other and influence each other, locally or globally.
- diverse: They possess entities of different types. (As opposed to variable: a difference in the value of an attribute.)
- adaptive/learning/intelligent: Their entities adapt their behavior to changes of other components in the system.
What is the difference between connectedness and interdependence?
Interdependence requires connectedness, but exerts different effect sizes depending on the degree of connectedness:
- High connectedness + low interdependence = no big effect
- High connectednesss + high interdependence = big effect
- Low connectedness + low interdependence = some effect
- Low connectedness + High interdepedence = moderate effect
What other features do complex systems exhibit?
- Selection: A process through which less fit or lower-performing entities are removed from the population.
- Robustness: The ability of a complex system to maintain functionality given a disturbance.
- Emergence: A higher-level phenomenon that arises from the micro-level interactions. (Weak: explicable; Strong: unexplicable yet, e.g. consciousness)
- Self-Organisation: A form of emergence in which the entities create a pattern or structure from the bottom-up.
- Self-Organised Criticality: A phenomenon in which interaction agents self-organise into states that can produce large events.
In contrast, what are complicated systems?
- connected
- interdependent
- diverse
- but not adaptive, not selective, not producing large events
What is it about the landscape analogy?
The landscape represents a problem or the fixed state of a complex system
- Encoding: The kind of landscape a problem or system produces, depends on the number of variables encoded. The length (and width) encode the variables. They describe the type of solution; the attribute(s) of the solution (e.g. shovel size)
- Height/Elevation represents the value of a potential solution. Local and global peaks represent (sub)optimal solutions.
- If the landscape becomes “bigger”, the number of possible solutions and non-solutions increases, and thus potentially the difficulty to solve the problem
What types of landscapes exist?
- simple landscapes: Neither interactions between actions of a single entity, nor interactions between multiple entities. They possess one global peak. (Example: optimisation of one design feature of one coal shovel.)
- rugged landscapes: Characterised by many possible combinations, and interactions within the actions of a single entity. They possess many local peaks with one (or multiple) global peaks. (Example: optimisation of all design features of one coal shovel; finding the optimal shovel from many good shovels.)
- dancing landscapes: Characterised by *externalities*, that is interactions between mulitple agents. They possess constantly changing local and global peaks. What is a solution today, might not be a solution tomorrow. They lie in the "interesting in-between": between a statistic regularity of high degrees to which the properties are pronounced and stasis of low degrees of property pronunciation. (Example: society)
How do we control complex systems?
- We build in some slack to increase robustness: Since the more a system is optimised, the more critical it becomes.
- We encourage diversity to balance exploration and exploitation, and prevent error.
- We perform safety measures for large events.
- We carefully define goals and incentives to control selection mechanisms:
- Connectedness: We create synergistic links and cut those that limits responsiveness.
What are the causes of diversity?
1. Diversity itself: the more you start with, the more combinations of the attributes of types are possible (positive feedback loop)
2. Weak selective pressure: The less types are removed from the population, the more types survive and reproduce.
3. Dancing landscapes and different landscapes:: An ever changing landscape produces multiple different landscapes that allow different types to solve problems.
What are the differences between evolutionary systems and creative systems?
1. Leap size: Evolution makes no leaps (all structures evolve slowly), while creative systems can jump forth.
2. Interim viability: Evolutionary outputs must be viable each step along the way to survive, while creative systems can produce outputs whose malfunction can be improved.
3. Representation: Evolution is fixed on genes, while creative systems can switch new ways of encoding.
November 26, 2019
This course touches on many different topics to try to explain and illustrate complexity (such as systems, networks, economics, evolution, and games). It's good in providing a framework around complexity. When the rules are known and the outcome is predictable, it's simple (like a game of tic-tac-toe). When the rules are not known and the outcome is unpredictable, it's chaos. Complexity is in the range between those two extremes. The subject of complexity isn't mature enough that it could be explained well. The lecturer uses a mountain as an analogy - seeking to reach the highest peak of the mountain. In a rugged landscape, you would identify the next higher peak and eventually you can reach the summit. However, in a "dancing" landscape (like a tectonic shift), the next higher peak could disappear and you find yourself in unfamiliar terrain so you must start the process again. Several analogies in the course isn't useful in explaining complexity.
November 6, 2019
A rare and distinctive example of taking something complex and making it completely understandable, without oversimplification. Moreover, it is a topic every curious person should know about. Many of the interesting phenomenon we encounter in real life, be it, consciousness, biology, climate change, or financial markets, can only be understood by seeing them as complex systems. I would strongly recommend this series of lectures to all my friends.
October 20, 2019
This is a set of interesting, entertaining, informative lectures on the science of complexity.
Much of the world we live in consists of complex systems, inherently changing, always in motion, or, as the author says, "dancing." They can't be controlled, but if we take the time to understand them, we can influence them. Properly applied, this could help prevent financial crashes, or prevent or contain epidemic. It can help design buildings better designed to enable people to evacuate safely in the event of a fire or other emergency.
Those are just the most obvious examples.
Page approaches the subject from a number of different angles. This is an introduction for an educated layperson like myself, not an in-depth course for those already familiar with the subject. I found it interesting, enjoyable, and informative all the way through, and am glad I listened to it.
Recommended.
I bought this audiobook.
Much of the world we live in consists of complex systems, inherently changing, always in motion, or, as the author says, "dancing." They can't be controlled, but if we take the time to understand them, we can influence them. Properly applied, this could help prevent financial crashes, or prevent or contain epidemic. It can help design buildings better designed to enable people to evacuate safely in the event of a fire or other emergency.
Those are just the most obvious examples.
Page approaches the subject from a number of different angles. This is an introduction for an educated layperson like myself, not an in-depth course for those already familiar with the subject. I found it interesting, enjoyable, and informative all the way through, and am glad I listened to it.
Recommended.
I bought this audiobook.
December 26, 2022
I expected they will talk about complexity but for me it felt the lectures are all over the place, crossing different areas from systems, nature, chemistry, philosophy… so I felt a bit lost and at the end didn’t get the point. I wish it was better explained or maybe better structured too. But supposedly we learn the basics of complex systems in these 12 lectures. Some insights I could catch are:
When we describe something as complex, we mean that it consists of interdependent, diverse entities, and we assume that those entities adapt - that they respond to their local and global environments.
There are 3 categories of landscape - simple (Mount Fuji - just one peak), rugged (many peaks and valleys), and dancing (qualities of interdependence and adaptability in complex systems create landscapes that are not just rugged but dancing). There are 2 types of peak: local and global. A local peak is a place on the landscape from which a step in any direction is a step down in elevation. A global peak is the highest of all of the local peaks of a given landscape. Most of the time the global peak is unique. Mount Fuji landscapes are single peaked; the local and global peak are by definition one and the same. Rugged landscapes have many local peaks, and it sometimes can be difficult to find the global peak. Dancing landscapes can have a single peak or multiple peaks, but the key feature is that those peaks change over time. In complex systems, agents adapt locally. If performance is considered as elevate on, then we can think of these agents as climbing hills. The local peaks are the best nearby options, whereas the global peaks are the best possible actions. Finding the highest point on a rugged landscape is not easy. The main reason is that the space of possibilities can be combinatorially huge. However, a large range of combinations is not the only thing necessary for a rugged landscape; the ingredients also have to interact. Once they do, the landscapes begin to have local peaks. These interactions occur within the choices of a single agent, as we can see when we consider the chain of effects created from the decision to remodel one aspect of a house. The simple rule is that the more interactions occur, the more rugged the landscape. The same logic holds for complex situations - how we encode them influences how quickly and effectively we can adapt.
Interdependence refers to whether other entities influence actions, connectedness refers to how many people a person is connected to. If a person is completely disconnected from everyone else, no one else can have any effect on that person’s actions. The result is not complex. If we hold the interdependency at a moderate level and raise the level of connectedness, we come up with some interesting results. At a somewhat low level of connectedness, equilibrium is established rather quickly. At a moderate level of connectedness, it can take a long while for equilibrium to be achieved. At a high level of connectedness, equilibrium is once again achieved quickly.
When we say diversity, we mean differences in types. We do not mean variations.
When we say a system is complex, we mean that it produces interesting nonperiodic patterns and emergent structures and functionalities. We have found that this complex state tends to lie in a region of moderate interdependence, moderate connectedness, some diversity, and some adaptation.
The two-armed bandit problem - imagine a slot machine with two levers, which offer the same payout but at a different rate. The problem is how to balance exploration (testing the rates of both levers) and exploitation (acting on a determination based on previous testing) to maximize return.
We saw how on a rugged landscape the balance between exploration and exploitation should end with almost complete exploitation. Hence we devise algorithms like simulated annealing, which start out exploring but end up exploiting. On dancing landscapes, agents can never stop exploring. This explains why we see complexity: Equilibrium allows for exploration, which stimulates dancing landscapes. Randomness is avoided because as exploration becomes prevalent, the value of exploitation increases. Thus individual agents balance the necessity to explore and exploit, producing complexity as a result. Complexity can thus be thought of as an emergent property. Emergence refers to the spontaneous creation of order and functionality from the bottom up. Systems with many parts can produce emergent phenomena that cannot be true of the parts themselves. A pool of water can be wet, but a single water molecule cannot.
A system self-organizes if the aggregation of individual actions produces an organized pattern at the macro level. A system is said to be critical if small events trigger large cascades. Therefore, self-organized criticality implies that systems self-organize so that what emerges is critical - it can produce big events. Central to our entire analysis is that the distribution of outcomes - the large events - depends on the complexity of the system. This idea that unpredictable events are not random but are the output of complexity represents a fundamental shift in perspective.
When we describe something as complex, we mean that it consists of interdependent, diverse entities, and we assume that those entities adapt - that they respond to their local and global environments.
There are 3 categories of landscape - simple (Mount Fuji - just one peak), rugged (many peaks and valleys), and dancing (qualities of interdependence and adaptability in complex systems create landscapes that are not just rugged but dancing). There are 2 types of peak: local and global. A local peak is a place on the landscape from which a step in any direction is a step down in elevation. A global peak is the highest of all of the local peaks of a given landscape. Most of the time the global peak is unique. Mount Fuji landscapes are single peaked; the local and global peak are by definition one and the same. Rugged landscapes have many local peaks, and it sometimes can be difficult to find the global peak. Dancing landscapes can have a single peak or multiple peaks, but the key feature is that those peaks change over time. In complex systems, agents adapt locally. If performance is considered as elevate on, then we can think of these agents as climbing hills. The local peaks are the best nearby options, whereas the global peaks are the best possible actions. Finding the highest point on a rugged landscape is not easy. The main reason is that the space of possibilities can be combinatorially huge. However, a large range of combinations is not the only thing necessary for a rugged landscape; the ingredients also have to interact. Once they do, the landscapes begin to have local peaks. These interactions occur within the choices of a single agent, as we can see when we consider the chain of effects created from the decision to remodel one aspect of a house. The simple rule is that the more interactions occur, the more rugged the landscape. The same logic holds for complex situations - how we encode them influences how quickly and effectively we can adapt.
Interdependence refers to whether other entities influence actions, connectedness refers to how many people a person is connected to. If a person is completely disconnected from everyone else, no one else can have any effect on that person’s actions. The result is not complex. If we hold the interdependency at a moderate level and raise the level of connectedness, we come up with some interesting results. At a somewhat low level of connectedness, equilibrium is established rather quickly. At a moderate level of connectedness, it can take a long while for equilibrium to be achieved. At a high level of connectedness, equilibrium is once again achieved quickly.
When we say diversity, we mean differences in types. We do not mean variations.
When we say a system is complex, we mean that it produces interesting nonperiodic patterns and emergent structures and functionalities. We have found that this complex state tends to lie in a region of moderate interdependence, moderate connectedness, some diversity, and some adaptation.
The two-armed bandit problem - imagine a slot machine with two levers, which offer the same payout but at a different rate. The problem is how to balance exploration (testing the rates of both levers) and exploitation (acting on a determination based on previous testing) to maximize return.
We saw how on a rugged landscape the balance between exploration and exploitation should end with almost complete exploitation. Hence we devise algorithms like simulated annealing, which start out exploring but end up exploiting. On dancing landscapes, agents can never stop exploring. This explains why we see complexity: Equilibrium allows for exploration, which stimulates dancing landscapes. Randomness is avoided because as exploration becomes prevalent, the value of exploitation increases. Thus individual agents balance the necessity to explore and exploit, producing complexity as a result. Complexity can thus be thought of as an emergent property. Emergence refers to the spontaneous creation of order and functionality from the bottom up. Systems with many parts can produce emergent phenomena that cannot be true of the parts themselves. A pool of water can be wet, but a single water molecule cannot.
A system self-organizes if the aggregation of individual actions produces an organized pattern at the macro level. A system is said to be critical if small events trigger large cascades. Therefore, self-organized criticality implies that systems self-organize so that what emerges is critical - it can produce big events. Central to our entire analysis is that the distribution of outcomes - the large events - depends on the complexity of the system. This idea that unpredictable events are not random but are the output of complexity represents a fundamental shift in perspective.
May 8, 2020
I really liked this course as an audiobook. It had a nice level to follow for an introductory level and some gave nice insights.
December 29, 2023
Any ethical implications at this time?…or that time?…Quick catch that butterfly and pin it to the cork board before his behavior becomes emergent!…Where’s the temp off to again?
Complexity is important to know about as a consumer of social media and other application designed to make us buy things. This course is actually good and gives away some actual secrets…Despite being ethically vacant and business terminology savvy, as expected.
This kind of complexity is like Keynesian or Marxist economics. If we all fully realized its importance we would be sunk. If we knew just how much complexity is used in creating algorithms that predict and create consumer behavior, we’d be devastated by its implications on personal and political events. We would all be studying it to be a responsible adult person in society. But like these other forms of economic study, which can be applied to causality and reductionism in all its domains, the most intelligent form of it will be kept secret by the same folks who provide top shelf funding to less practical optics based courses.
We were busy with Reaganomics for more than a generation. Trying to figure out faster and faster how trickle down economics might just work. Thinking that supply and demand still existed, and not just overproduction and force feeding. Thinking that a measure of time could still be quantified like the industrial era.
Even if we took Reaganomics seriously we would have to realize that a free economy in the marketplace somehow meant, in practical terms, that a public federal government would regulate everything to ensure the most successful businesses and war machines continue to be successful. Since public and private became the same thing right around then.
Reagan said one thing and then used government to do the other thing. So basically it ended up the opposite of a free market. Not super complex, just saying one thing and meaning the complete opposite thing when using the science as an apparatus of power…which mainly worked because he lived in the era when the richest folks finally achieved oligarchic level dividends.
Once Marxism was destroyed with the false belief that it was the same thing as totalitarian communism, they began educating people with useless academic propaganda and getting us all to pay for it. And it’s supposed to be the bachelors in arts that’s the useless one! Oh life, in all your complexity, we are so indebted to understanding you…at a later time.
Complexity is important to know about as a consumer of social media and other application designed to make us buy things. This course is actually good and gives away some actual secrets…Despite being ethically vacant and business terminology savvy, as expected.
This kind of complexity is like Keynesian or Marxist economics. If we all fully realized its importance we would be sunk. If we knew just how much complexity is used in creating algorithms that predict and create consumer behavior, we’d be devastated by its implications on personal and political events. We would all be studying it to be a responsible adult person in society. But like these other forms of economic study, which can be applied to causality and reductionism in all its domains, the most intelligent form of it will be kept secret by the same folks who provide top shelf funding to less practical optics based courses.
We were busy with Reaganomics for more than a generation. Trying to figure out faster and faster how trickle down economics might just work. Thinking that supply and demand still existed, and not just overproduction and force feeding. Thinking that a measure of time could still be quantified like the industrial era.
Even if we took Reaganomics seriously we would have to realize that a free economy in the marketplace somehow meant, in practical terms, that a public federal government would regulate everything to ensure the most successful businesses and war machines continue to be successful. Since public and private became the same thing right around then.
Reagan said one thing and then used government to do the other thing. So basically it ended up the opposite of a free market. Not super complex, just saying one thing and meaning the complete opposite thing when using the science as an apparatus of power…which mainly worked because he lived in the era when the richest folks finally achieved oligarchic level dividends.
Once Marxism was destroyed with the false belief that it was the same thing as totalitarian communism, they began educating people with useless academic propaganda and getting us all to pay for it. And it’s supposed to be the bachelors in arts that’s the useless one! Oh life, in all your complexity, we are so indebted to understanding you…at a later time.
May 19, 2022
When the landscape dances, we must adapt.
Complexity explained easily. I just wished I had read it before. Plenty of down to earth examples that make a complicated topic digestible. I also loved the summaries at the end of each chapter. So many takeaways from this book. Here’s a few from my notes on the “harnessing complexity” chapter:
- diversity is beneficial to innovation and robustness. But if agents are too diverse we need to balance diversity or we will have selection driving exploitation at the expense of exploration.
- Avoid dominant logics that leads to group thinking.
- Keep on experimenting, like evolution does. Don’t be blinded by dominant logics. Mindlessly try every possible combination.
- Interdependencies: don’t push the system to a critical state by improving the efficiencies. Build some slack so you don’t have cascading failures sparked by a small failure. Leave room for slippage. Allowing slippage also increases diversity, therefore innovation. Don’t get too caught up on little efficiency gains.
- Keep an eye on the tails.
- Connections: Sever unnecessary connections that limits innovation responsiveness , build connections that produce synergies. Synergistic links exploit diversity and positive interdependencies.
- Set the right incentives with the aim of bridging links between domains.
Complexity explained easily. I just wished I had read it before. Plenty of down to earth examples that make a complicated topic digestible. I also loved the summaries at the end of each chapter. So many takeaways from this book. Here’s a few from my notes on the “harnessing complexity” chapter:
- diversity is beneficial to innovation and robustness. But if agents are too diverse we need to balance diversity or we will have selection driving exploitation at the expense of exploration.
- Avoid dominant logics that leads to group thinking.
- Keep on experimenting, like evolution does. Don’t be blinded by dominant logics. Mindlessly try every possible combination.
- Interdependencies: don’t push the system to a critical state by improving the efficiencies. Build some slack so you don’t have cascading failures sparked by a small failure. Leave room for slippage. Allowing slippage also increases diversity, therefore innovation. Don’t get too caught up on little efficiency gains.
- Keep an eye on the tails.
- Connections: Sever unnecessary connections that limits innovation responsiveness , build connections that produce synergies. Synergistic links exploit diversity and positive interdependencies.
- Set the right incentives with the aim of bridging links between domains.
December 16, 2024
the topic is specifically interesting for building predictive models and the course is definitely better than others from the Great Courses collection, that said, i noticed that most of the research done in the field, at least based on this, looks at evolution, as a base for understanding complexity, a lot, and i would have liked for this to talk more about abstract logical concept theory, to not fall into the trap of over-optimising the research function and end up with the theoretical model overfitting.
the knocking-out interconnected nodes chapter gave me flashbacks and was particularly interesting, as a software engineer, it's an important part of testing systems and making sure they're robust and keep performing under stress...a concept at the base of any big tech company in which they control their own servers and data bases offering services, especially highly used ones...Amazon is making a killing at this these days, it would be interesting to bring more into the theory, and examples, from the computer science field. machine learning has also probably produced more interesting theories in this area, complexity in general, especially lately...i felt it this case, with the nodes, it contained the same things i learned about more than a decade ago in school.
the knocking-out interconnected nodes chapter gave me flashbacks and was particularly interesting, as a software engineer, it's an important part of testing systems and making sure they're robust and keep performing under stress...a concept at the base of any big tech company in which they control their own servers and data bases offering services, especially highly used ones...Amazon is making a killing at this these days, it would be interesting to bring more into the theory, and examples, from the computer science field. machine learning has also probably produced more interesting theories in this area, complexity in general, especially lately...i felt it this case, with the nodes, it contained the same things i learned about more than a decade ago in school.
May 29, 2021
This Great Course reminded me of "Freakonomics" or a Malcolm Gladwell book rendered more rigorous: it discusses the unlikely connections that create complex systems in which there is no true top-down organization, but rather a series of subtle interconnections. Early lectures allude to Chaos Theory to explain these connections, and particularly the butterfly flapping its wings that can later lead to a hurricane on the other side of the world. Cultures, economies, climate, politics, marketing, and physiology would all fit into this category. They would all also "dance," as Page would say: that is, as one variable changes, all of the other variables simultaneously adjust to account for the change. It seems to me that this concept is lacking in much of the Western worldview on nearly all of those topics; we try to hit every problem head-on with a sledgehammer, but fail to account for all of the other dancing variables that will compensate for the blow. In medicine, these are called "side effects"--yet we persist in suppressing symptoms. So too in economics, in politics, etc, addressing the symptom will merely create new problems, sometimes worse than the original "disease."
Displaying 1 - 30 of 107 reviews