What do you think?
Rate this book
Social Physics: How Good Ideas Spread— The Lessons from a New Science
From one of the world’s leading data scientists, a landmark tour ofthe new science of idea flow, offering revolutionary insights into the mysteries of collective intelligence and social influence
If the Big Data revolution has a presiding genius, it is MIT’s Alex "Sandy" Pentland. Over years of groundbreaking experiments, he has distilled remarkable discoveries significant enough to become the bedrock of a whole new scientific field: social physics. Humans have more in common with bees than we like to admit: We’re social creatures first and foremost. Our most important habits of action—and most basic notions of common sense—are wired into us through our coordination in social groups. Social physics is about idea flow, the way human social networks spread ideas and transform those ideas into behaviors.
Thanks to the millions of digital bread crumbs people leave behind via smartphones, GPS devices, and the Internet, the amount of new information we have about human activity is truly profound. Until now, sociologists have depended on limited data sets and surveys that tell us how people say they think and behave, rather than what they actually do. As a result, we’ve been stuck with the same stale social structures—classes, markets—and a focus on individual actors, data snapshots, and steady states. Pentland shows that, in fact, humans respond much more powerfully to social incentives that involve rewarding others and strengthening the ties that bind than incentives that involve only their own economic self-interest.
Pentland and his teams have found that they can study patterns of information exchange in a social network without any knowledge of the actual content of the information and predict with stunning accuracy how productive and effective that network is, whether it’s a business or an entire city. We can maximize a group’s collective intelligence to improve performance and use social incentives to create new organizations and guide them through disruptive change in a way that maximizes the good. At every level of interaction, from small groups to large cities, social networks can be tuned to increase exploration and engagement, thus vastly improving idea flow.
Social Physics will change the way we think about how we learn and how our social groups work—and can be made to work better, at every level of society. Pentland leads readers to the edge of the most important revolution in the study of social behavior in a generation, an entirely new way to look at life itself.
If the Big Data revolution has a presiding genius, it is MIT’s Alex "Sandy" Pentland. Over years of groundbreaking experiments, he has distilled remarkable discoveries significant enough to become the bedrock of a whole new scientific field: social physics. Humans have more in common with bees than we like to admit: We’re social creatures first and foremost. Our most important habits of action—and most basic notions of common sense—are wired into us through our coordination in social groups. Social physics is about idea flow, the way human social networks spread ideas and transform those ideas into behaviors.
Thanks to the millions of digital bread crumbs people leave behind via smartphones, GPS devices, and the Internet, the amount of new information we have about human activity is truly profound. Until now, sociologists have depended on limited data sets and surveys that tell us how people say they think and behave, rather than what they actually do. As a result, we’ve been stuck with the same stale social structures—classes, markets—and a focus on individual actors, data snapshots, and steady states. Pentland shows that, in fact, humans respond much more powerfully to social incentives that involve rewarding others and strengthening the ties that bind than incentives that involve only their own economic self-interest.
Pentland and his teams have found that they can study patterns of information exchange in a social network without any knowledge of the actual content of the information and predict with stunning accuracy how productive and effective that network is, whether it’s a business or an entire city. We can maximize a group’s collective intelligence to improve performance and use social incentives to create new organizations and guide them through disruptive change in a way that maximizes the good. At every level of interaction, from small groups to large cities, social networks can be tuned to increase exploration and engagement, thus vastly improving idea flow.
Social Physics will change the way we think about how we learn and how our social groups work—and can be made to work better, at every level of society. Pentland leads readers to the edge of the most important revolution in the study of social behavior in a generation, an entirely new way to look at life itself.
320 pages, Hardcover
First published January 1, 2014
339 people are currently reading
3514 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 184 reviews
June 8, 2020
This very well written book contains great ideas that not only are interesting in theory, but very promising if applied to society, as the experiments the author and his team seem to demonstrate. However, this book has a huge problem: it reduces a big picture to the authors particular research, not giving credit nor connecting with multiple different approaches to the same problems. A naïve reader will get the impression that the author and his team are the pioneers or, at minimal, the leading researchers, of several fields such as cultural analytics, memetics, ideas spread, social networks, open data legislation, mathematics of cooperation and collective intelligence.
As another commentator put it, the book feels like an academic resumé, and I'd add it also feels like a brochure for the companies the author has found (along with his students). This academic nepotism is bluntly paradoxical, being a book about cooperation, ideas flow, social intelligence and the importance of diversity and exploration of external ideas.
The book even fails to mention that "social physics" is a term famously used by August Comte, and previously by other sociology pioneers. (I think Pentland believes he coined the term).
It feels sad to give 2 stars to a book that has great contents, but it would be very easy for the author to have expanded the already interesting ideas, connecting them with the work of many other great thinkers and well stablished approaches; the book would gain so much.
As another commentator put it, the book feels like an academic resumé, and I'd add it also feels like a brochure for the companies the author has found (along with his students). This academic nepotism is bluntly paradoxical, being a book about cooperation, ideas flow, social intelligence and the importance of diversity and exploration of external ideas.
The book even fails to mention that "social physics" is a term famously used by August Comte, and previously by other sociology pioneers. (I think Pentland believes he coined the term).
It feels sad to give 2 stars to a book that has great contents, but it would be very easy for the author to have expanded the already interesting ideas, connecting them with the work of many other great thinkers and well stablished approaches; the book would gain so much.
February 22, 2014
Could be an interesting topic but the writing was dry and the author talked about himself all the time.
February 8, 2024
I am ashamed of this book and the academy. The next time any academic thinks bad about Malcolm Gladwell, please think twice, read this and find out why most academics shouldn't publish books for the public.
Alex Pentland is a high praised academic from MIT in the domain of Big Data, but this book is not about research, neither about promoting research. This book is more about his achievements memoirs. Do we really want to know about where his publications were accepted, and were he works, or should we be reading about the intricacies of the research its explanations and interpretations?
For someone advocating through the entire book the collaborative side of humanity, it's really painful seeing that the way used the present all the research gives the idea of a "one-man show".
In the end this book was not written for the public, neither for academics, this book was written as a promotion pamphlet to attract venture capitalists to invest in his laboratory. In part it shows what is happening to the academy and the universities, that are being transformed in comercial corporations, forgetting knowledge, forgetting its origins and purposes.
About the ideas, called novel in this book, I'll write soon in my blog.
Blog review in portuguese: http://virtual-illusion.blogspot.pt/2...
Alex Pentland is a high praised academic from MIT in the domain of Big Data, but this book is not about research, neither about promoting research. This book is more about his achievements memoirs. Do we really want to know about where his publications were accepted, and were he works, or should we be reading about the intricacies of the research its explanations and interpretations?
For someone advocating through the entire book the collaborative side of humanity, it's really painful seeing that the way used the present all the research gives the idea of a "one-man show".
In the end this book was not written for the public, neither for academics, this book was written as a promotion pamphlet to attract venture capitalists to invest in his laboratory. In part it shows what is happening to the academy and the universities, that are being transformed in comercial corporations, forgetting knowledge, forgetting its origins and purposes.
About the ideas, called novel in this book, I'll write soon in my blog.
Blog review in portuguese: http://virtual-illusion.blogspot.pt/2...
March 9, 2014
Pentland and his research group have appeared to have discovered a simple model of human behavior with great predictive power. By snooping on people's cell phones, they can reduce typical human interactions into a set of interacting finite state machines, and by noticing just how regular those behavioral patterns are they think they understand ideas. Quite evidently the populations they study live routine, predictable lives. (Perhaps we all do.)
My biggest criticism (& thus the non-recommendation for the book) is that the technologies for making these monitoring of human behavior that Pentland describes are dehumanizing and a grave insult to personal privacy. The vague promises of a personal data store are unrealistic. If these plans pan out, we will always be watched over by machines that seek only to predict our patterns of behavior and exploit them.
Every once in a while the author describes unexpected behavior by individuals thus modeled, and betrays an element of surprise that we might step out of our everyday paths into something his system doesn't contain. It's a glimmer of hope in a dystopian world that we might surprise our ever-present overseers and do something that their social physics does not anticipate.
My biggest criticism (& thus the non-recommendation for the book) is that the technologies for making these monitoring of human behavior that Pentland describes are dehumanizing and a grave insult to personal privacy. The vague promises of a personal data store are unrealistic. If these plans pan out, we will always be watched over by machines that seek only to predict our patterns of behavior and exploit them.
Every once in a while the author describes unexpected behavior by individuals thus modeled, and betrays an element of surprise that we might step out of our everyday paths into something his system doesn't contain. It's a glimmer of hope in a dystopian world that we might surprise our ever-present overseers and do something that their social physics does not anticipate.
February 16, 2014
This book reads like a resume from the author. Through out the book the author claims he and his team made an assortment of discoveries when in all actuality his team merely confirmed experiments conducted by other scientists. Also, the fact that the author makes no mention of memetics shows his naivety in the field, if his repeated extrapolations weren't enough. If this subject interests you try reading "The Half-Life of Facts by Samuel Arbesman" instead.
February 25, 2014
*A full executive summary of this book is available here: http://newbooksinbrief.com/2014/02/25...
The main argument: The sciences that focus on human behavior, meaning the social sciences, have traditionally relied mainly on surveys and lab experiments in their investigations. While valuable to a degree, these sources of evidence do have their shortcomings. Most significantly, surveys offer but indirect evidence of human behavior (and can also be compromised by deception and self-deception); while lab experiments tend to be somewhat artificial, and fail to capture the complexities of real life.
Recently, however, new digital technology has opened up a whole new way to study human behavior. This proves to be the case since mobile devices and sensors of all kinds are now able to record a dizzying array of human activity—everything from where we go, to what we buy, to whom we interact with and for how long, to our body language, and even our moods etc. When placed in the hands of social scientists these new sources of information can prove very valuable (and are far preferable than either surveys or lab experiments); for they allow scientists to study us in our natural environments—out in the real world—and they also allow scientists to study what we actually do, rather than what we say (which are sometimes quite different).
The method of investigating human behavior in our natural environments using digital technology has come to be called reality mining, and it is revolutionizing the social sciences.
One of the pioneers and leaders in the field of reality mining is Alex Pentland, a researcher out of MIT. Pentland’s main field of interest is using reality mining to explore the properties and patterns of interactions between people—what he calls social physics. Specifically, Pentland uses reality mining to investigate the social physics in a wide range of groups and situations, from social and peer groups; to social media platforms; to institutional settings such as schools and businesses; to even whole cities. And in his new book Social Physics: How Good Ideas Spread—The Lessons from a New Science Pentland takes time out to catch us up on his findings.
One of Pentlands’s main findings thus far has to do with the importance of social interaction in influencing our behavior. Indeed, Pentland has found that much of our behavior is dominated by the influence of our close relations and the peer groups we are embedded in—everything from our diet and body weight to our political opinions and all things in between.
The influence of our social world is so great, in fact, that Pentland argues it is much more appropriate to think of ourselves as group-oriented than self-directed. This is important because Western society as a whole tends to take the opposite view. The result is that many of our policies and institutions are ill-fitted to our true nature—which leads to less than desirable outcomes. Thankfully, Pentland does offer some advice with regards to how we can re-design our policies and institutions in a way that better accommodates our nature.
A second of Pentland’s main findings has to do with how ideas and behavior spread through human interactions and groups—and also, and even more important, what kinds of interactions produce the best results in terms of generating the most creative and productive ideas.
Specifically, Pentland has found that the most creative and productive groups tend to have something very important in common: the group members have numerous interactions with highly diverse people outside of the group, and the group members are also highly connected to one another.
In terms of explaining why this pattern works best Pentland argues that the interactions outside of the group are important in becoming familiar with many different types of ideas, while the interactions within the group function to winnow out what are the best ideas, and also help build common norms of behavior and trust that allow the group to work well and cooperatively together.
I was happy to get the opportunity to learn about a very new and promising science from one of its leading practitioners. Many of the ‘living lab’ experiments outlined in the book are very interesting and I certainly learned a lot. My only complaints are that the book does have a fair bit of repetition and jumps around some, so I question the writing and organization a bit. All in all, though, a very good and interesting read about a new field that we are sure to hear more from moving forward. A full executive summary of the book is available here: http://newbooksinbrief.com/2014/02/25...
The main argument: The sciences that focus on human behavior, meaning the social sciences, have traditionally relied mainly on surveys and lab experiments in their investigations. While valuable to a degree, these sources of evidence do have their shortcomings. Most significantly, surveys offer but indirect evidence of human behavior (and can also be compromised by deception and self-deception); while lab experiments tend to be somewhat artificial, and fail to capture the complexities of real life.
Recently, however, new digital technology has opened up a whole new way to study human behavior. This proves to be the case since mobile devices and sensors of all kinds are now able to record a dizzying array of human activity—everything from where we go, to what we buy, to whom we interact with and for how long, to our body language, and even our moods etc. When placed in the hands of social scientists these new sources of information can prove very valuable (and are far preferable than either surveys or lab experiments); for they allow scientists to study us in our natural environments—out in the real world—and they also allow scientists to study what we actually do, rather than what we say (which are sometimes quite different).
The method of investigating human behavior in our natural environments using digital technology has come to be called reality mining, and it is revolutionizing the social sciences.
One of the pioneers and leaders in the field of reality mining is Alex Pentland, a researcher out of MIT. Pentland’s main field of interest is using reality mining to explore the properties and patterns of interactions between people—what he calls social physics. Specifically, Pentland uses reality mining to investigate the social physics in a wide range of groups and situations, from social and peer groups; to social media platforms; to institutional settings such as schools and businesses; to even whole cities. And in his new book Social Physics: How Good Ideas Spread—The Lessons from a New Science Pentland takes time out to catch us up on his findings.
One of Pentlands’s main findings thus far has to do with the importance of social interaction in influencing our behavior. Indeed, Pentland has found that much of our behavior is dominated by the influence of our close relations and the peer groups we are embedded in—everything from our diet and body weight to our political opinions and all things in between.
The influence of our social world is so great, in fact, that Pentland argues it is much more appropriate to think of ourselves as group-oriented than self-directed. This is important because Western society as a whole tends to take the opposite view. The result is that many of our policies and institutions are ill-fitted to our true nature—which leads to less than desirable outcomes. Thankfully, Pentland does offer some advice with regards to how we can re-design our policies and institutions in a way that better accommodates our nature.
A second of Pentland’s main findings has to do with how ideas and behavior spread through human interactions and groups—and also, and even more important, what kinds of interactions produce the best results in terms of generating the most creative and productive ideas.
Specifically, Pentland has found that the most creative and productive groups tend to have something very important in common: the group members have numerous interactions with highly diverse people outside of the group, and the group members are also highly connected to one another.
In terms of explaining why this pattern works best Pentland argues that the interactions outside of the group are important in becoming familiar with many different types of ideas, while the interactions within the group function to winnow out what are the best ideas, and also help build common norms of behavior and trust that allow the group to work well and cooperatively together.
I was happy to get the opportunity to learn about a very new and promising science from one of its leading practitioners. Many of the ‘living lab’ experiments outlined in the book are very interesting and I certainly learned a lot. My only complaints are that the book does have a fair bit of repetition and jumps around some, so I question the writing and organization a bit. All in all, though, a very good and interesting read about a new field that we are sure to hear more from moving forward. A full executive summary of the book is available here: http://newbooksinbrief.com/2014/02/25...
November 29, 2020
This is not some pop-science book by a run-of-the-mill academic: Alex Pentland is one of the most powerful scientists in the world. His projects outlined here are not just experiments: they are prototypes exploring what a surveillance-based society might look like. Which is why Pentland's avoidance of any discussion of *power* in the book - who has it, who wields it and for what end - feels less ignorance than dissemblance.
There is a lot here to agree with - Pentland's explication of how systemic social dynamics can create or ruin a company's output, for example, should be compulsory reading for managements obsessed with time metrics and personality typing. I tend to agree with the analysis of the role of collaboration in driving human evolution. His argument that exploration and novel exposure drive innovation, and his enthusiasm for structuring networks and systems to support this is compelling.
In part, I think Pentland's ideas are dangerous because of what he's right about: exactly what is needed to manipulate people. And the dangerous part is not so much what he is wrong about, but what he glosses over and ignores: the question of who benefits from this manipulation, who has the power to carry it out and how this is governed.
It is notable that Pentland does not discuss the issues around data analysis and equality. There is no mention of how this technology is being used to deepen racist housing and schooling segregation, or to 'predict' criminality based on race, or reduce loan limits for women. In recent years, as the evidence for drastically worsening racial injustice from large data systems has increased, Pentland has been involved in a few projects around "fairness" in tech, including one which facilitates a tool allowing decision makers to weigh how 'unfair' the system is vs how much it might cost to make it 'fairer'.
He diverts from this by much content around individual privacy and control over data. How this sits with his work in organisations - in which all employees must submit to constant monitoring via a biometric badge and total system surveillance is not specifically discussed. That Pentland's findings support more engaging and social workplace is no doubt some incentive to the surveillance, but the end objective is not to improve their health and wellbeing but to improve productivity for the corporations employing them.
Similarly, Pentland's smart cities experiments can be sold to participants as improving traffic flow, better targeting of health services and disease control. In reality, of course, this technology is wielded almost entirely to get people to buy things when they are most vulnerable to suggestion. Our travel routes are not designed to maximise our pleasure and relaxation but to ensure we notice that coffee shop we might love, or that pet shop with the puppy we can't resist. Our vulnerabilities are on sale to the highest bidder. The overall message: that the way we construct our societies determines what we can achieve, I find very attractive.
Pentland, of course, doesn't bring this up at all. Instead, he waxes lyrical about the potential for constant smart surveillance for social improvements. We can know, he enthuses, where people who are going to get diabetes visit. Now I'm not a data scientist or a dietician, but even I know that fast food outlets are a major contributor to poor health outcomes. This isn't exactly a secret - our problems aren't about knowledge, they are about power and the power that the profit holds. Pentland argues explicitly at several points that markets have become divorced from social good because of a lack of knowledge. Most cheekily, he tries to imply the industrial revolution was hard on the poor because of a lack of cross-class socialising (industrial barons would have given up all that power and wealth if they just understood their workers' grief, physical pains, and exhaustion and starvation better.)
Never is this more hubristic than in Pentland's claim that data will enable us to avoid or minimise pandemics. Of course, I get to be late 2020 hysterically hindsighty about this but turns out Pentland is right about what data *could* do, but totally wrong about our ability to use it for public good. 'Contact tracing' is now ordinary language and data modellers have had a hell of a year, but in the heartland of the market surveillance model Pentland advocates, it has had minimal impact on public policy, and hundreds of thousands have died. As I write this, millions of Americans gather for Thanksgiving in open violation of recommended protections, in a push largely motivated by ensuring continuing commodity consumption driven by a market economy in direct conflict with public safety.
I'm not much of a libertarian. I have little time for the idea of some pure individual unshaped or unmanipulated by forces around us. I don't hate the 'social physics' discipline because it snoops on us, or is designed to alter our behaviour. Human cognition is socially constructed - we are shaped by our society all the time. But I certainly don't believe it is a new organising model either. It is just a tool - an extraordinarily powerful one. The question is who wields it, and at what cost?
There is a lot here to agree with - Pentland's explication of how systemic social dynamics can create or ruin a company's output, for example, should be compulsory reading for managements obsessed with time metrics and personality typing. I tend to agree with the analysis of the role of collaboration in driving human evolution. His argument that exploration and novel exposure drive innovation, and his enthusiasm for structuring networks and systems to support this is compelling.
In part, I think Pentland's ideas are dangerous because of what he's right about: exactly what is needed to manipulate people. And the dangerous part is not so much what he is wrong about, but what he glosses over and ignores: the question of who benefits from this manipulation, who has the power to carry it out and how this is governed.
It is notable that Pentland does not discuss the issues around data analysis and equality. There is no mention of how this technology is being used to deepen racist housing and schooling segregation, or to 'predict' criminality based on race, or reduce loan limits for women. In recent years, as the evidence for drastically worsening racial injustice from large data systems has increased, Pentland has been involved in a few projects around "fairness" in tech, including one which facilitates a tool allowing decision makers to weigh how 'unfair' the system is vs how much it might cost to make it 'fairer'.
He diverts from this by much content around individual privacy and control over data. How this sits with his work in organisations - in which all employees must submit to constant monitoring via a biometric badge and total system surveillance is not specifically discussed. That Pentland's findings support more engaging and social workplace is no doubt some incentive to the surveillance, but the end objective is not to improve their health and wellbeing but to improve productivity for the corporations employing them.
Similarly, Pentland's smart cities experiments can be sold to participants as improving traffic flow, better targeting of health services and disease control. In reality, of course, this technology is wielded almost entirely to get people to buy things when they are most vulnerable to suggestion. Our travel routes are not designed to maximise our pleasure and relaxation but to ensure we notice that coffee shop we might love, or that pet shop with the puppy we can't resist. Our vulnerabilities are on sale to the highest bidder. The overall message: that the way we construct our societies determines what we can achieve, I find very attractive.
Pentland, of course, doesn't bring this up at all. Instead, he waxes lyrical about the potential for constant smart surveillance for social improvements. We can know, he enthuses, where people who are going to get diabetes visit. Now I'm not a data scientist or a dietician, but even I know that fast food outlets are a major contributor to poor health outcomes. This isn't exactly a secret - our problems aren't about knowledge, they are about power and the power that the profit holds. Pentland argues explicitly at several points that markets have become divorced from social good because of a lack of knowledge. Most cheekily, he tries to imply the industrial revolution was hard on the poor because of a lack of cross-class socialising (industrial barons would have given up all that power and wealth if they just understood their workers' grief, physical pains, and exhaustion and starvation better.)
Never is this more hubristic than in Pentland's claim that data will enable us to avoid or minimise pandemics. Of course, I get to be late 2020 hysterically hindsighty about this but turns out Pentland is right about what data *could* do, but totally wrong about our ability to use it for public good. 'Contact tracing' is now ordinary language and data modellers have had a hell of a year, but in the heartland of the market surveillance model Pentland advocates, it has had minimal impact on public policy, and hundreds of thousands have died. As I write this, millions of Americans gather for Thanksgiving in open violation of recommended protections, in a push largely motivated by ensuring continuing commodity consumption driven by a market economy in direct conflict with public safety.
I'm not much of a libertarian. I have little time for the idea of some pure individual unshaped or unmanipulated by forces around us. I don't hate the 'social physics' discipline because it snoops on us, or is designed to alter our behaviour. Human cognition is socially constructed - we are shaped by our society all the time. But I certainly don't believe it is a new organising model either. It is just a tool - an extraordinarily powerful one. The question is who wields it, and at what cost?
May 28, 2014
Conteúdo legal, boas ideias e bastante pesquisa de ponta por trás desse livro. O que me deixou mais surpreso com a forma como foi escrito. Muita auto-promoção do autor, muito "isso eu e fulano investigamos" e quase nada sobre qualquer outra pesquisa na área. Vários trechos se resumem a repetir a mesma ideia de que interações são importantes e como o autor fundou uma empresa para explorar aqueles dados. Vale por algumas explicações, se você tiver a paciência de aguentar a auto-promoção excessiva.
April 26, 2014
I went through this book in one sitting. Reminds me of Rational Choice Theory but with a layer of real-time data update and Data-driven-decision-making.
I plan to go back and read "Honest Signals" also by the same author. I had meant to. I'm more motivated now.
I plan to go back and read "Honest Signals" also by the same author. I had meant to. I'm more motivated now.
January 21, 2021
I feel a bit weird after finishing this book... the topic of Social Network Analysis is something that interests me very much, but this book didn't provide me a lot of new in depth knowledge. Instead, I got to learn about all the papers that Alex Pentland has published (and where), all the companies he has co-founded, all the times he was involved in international trans-institutional policy making etcetera etcetera. So it did a splendid job at promoting Alex Pentland himself.
His point was clear: instead of analysing societies through the lens of markets or classes, we should preform analysis on social interactions between individuals. He promotes the use of computational social science/big data analysis to get insights in how cultures develop in terms of engagement and exploration. In his eyes, social physics can serve as the foundation of the solution to virtually all the problems of the world.
Although I agree with him that SNA is a very promising field of science, I wish he would've chosen to present it from a broader prospectie than just from his MIT lab (which he founded) and where he works (with his PhD-students (with whom he co-founded a lot of companies)). His concerns about privacy are also a bit vague tbh. He keeps referring to the last chapter of the book: 'I'll explain everything about privacy there', but it turned out the be just the final paragraphs of the final chapter, in which he remains quite vague.
Anyways, his explanations of SNA are alright. It annoyed me that he made it look like he invented all that stuff, without mentioning legends in the field such as Granovetter and Burt (he briefly mentioned Burt), or the small-world-theory. He did present some cool cases that he worked on, but I def got some narcissistic vibe of it.
Would I recommend it? Good question. I think yes maybe, just for to get acquainted with the cool cases that he worked.
His point was clear: instead of analysing societies through the lens of markets or classes, we should preform analysis on social interactions between individuals. He promotes the use of computational social science/big data analysis to get insights in how cultures develop in terms of engagement and exploration. In his eyes, social physics can serve as the foundation of the solution to virtually all the problems of the world.
Although I agree with him that SNA is a very promising field of science, I wish he would've chosen to present it from a broader prospectie than just from his MIT lab (which he founded) and where he works (with his PhD-students (with whom he co-founded a lot of companies)). His concerns about privacy are also a bit vague tbh. He keeps referring to the last chapter of the book: 'I'll explain everything about privacy there', but it turned out the be just the final paragraphs of the final chapter, in which he remains quite vague.
Anyways, his explanations of SNA are alright. It annoyed me that he made it look like he invented all that stuff, without mentioning legends in the field such as Granovetter and Burt (he briefly mentioned Burt), or the small-world-theory. He did present some cool cases that he worked on, but I def got some narcissistic vibe of it.
Would I recommend it? Good question. I think yes maybe, just for to get acquainted with the cool cases that he worked.
January 2, 2019
A high 3, around 4 stars!
I did it! I finally finished this book and just in time for school to start tomorrow! A little bit of stress was felt while reading the last 60 pages of this book, since I was trying to get it finished by the start of school, but I actually did enjoy this book overall. I did also feel a little bit rushed and lacked focus, all is good now.
I almost didn't start this book for many reasons, but I am glad that I did. Alex Pentland is smart and insightful about things and I have learned so much, my mind has opened to a new perspective, just from reading this book.
Before this book, I didn't know what social physics was or how it could even benefit society. I am a person interested in sociology and the ideas that Alex presented in this book were very good and made so much sense that I was surprised that I haven't heard of them. Even though, social physics sounds like a scary topic and like a topic that would be dense and hard to understand, Alex writes in a way that makes these BIG SCARY TOPICS more understandable and connectable to everyday people. I didn't feel talked-down to, which I really appreciated and it made me get into the book easier, the flow of reading it and taking it in, since I could understand what he was trying to get across.
With this, it was a smart idea for Alex to put most of the math in an appendix, I found the concepts easier to understand without the math staring me down and intimidating me.
I felt like some of the book was repetitive, but I did like how he would review things or preview things later to come. It really helped me not lose myself and to keep track of Alex's ideas. He did most of the studies in this book and you could tell how intelligent and well-knowledgeable to talk about all that he talked about.
The math appendix was kind of boring to me and I didn't really enjoy it, except for the more analytic wordy sections. I was glad that it was short. It did give me a headache. I think some people might enjoy it, but I as a person that is bad at math didn't. I found the other appendixes really interesting, got to see how things that Alex described in his book worked.
Some people might find this book boring, but I was not one of them. I felt like my mind was expanding and that I could take many of the things from this book and apply them to my life. I really thought that this book was interesting. It didn't wow me, but I liked it and enjoyed it.
I did it! I finally finished this book and just in time for school to start tomorrow! A little bit of stress was felt while reading the last 60 pages of this book, since I was trying to get it finished by the start of school, but I actually did enjoy this book overall. I did also feel a little bit rushed and lacked focus, all is good now.
I almost didn't start this book for many reasons, but I am glad that I did. Alex Pentland is smart and insightful about things and I have learned so much, my mind has opened to a new perspective, just from reading this book.
Before this book, I didn't know what social physics was or how it could even benefit society. I am a person interested in sociology and the ideas that Alex presented in this book were very good and made so much sense that I was surprised that I haven't heard of them. Even though, social physics sounds like a scary topic and like a topic that would be dense and hard to understand, Alex writes in a way that makes these BIG SCARY TOPICS more understandable and connectable to everyday people. I didn't feel talked-down to, which I really appreciated and it made me get into the book easier, the flow of reading it and taking it in, since I could understand what he was trying to get across.
With this, it was a smart idea for Alex to put most of the math in an appendix, I found the concepts easier to understand without the math staring me down and intimidating me.
I felt like some of the book was repetitive, but I did like how he would review things or preview things later to come. It really helped me not lose myself and to keep track of Alex's ideas. He did most of the studies in this book and you could tell how intelligent and well-knowledgeable to talk about all that he talked about.
The math appendix was kind of boring to me and I didn't really enjoy it, except for the more analytic wordy sections. I was glad that it was short. It did give me a headache. I think some people might enjoy it, but I as a person that is bad at math didn't. I found the other appendixes really interesting, got to see how things that Alex described in his book worked.
Some people might find this book boring, but I was not one of them. I felt like my mind was expanding and that I could take many of the things from this book and apply them to my life. I really thought that this book was interesting. It didn't wow me, but I liked it and enjoyed it.
February 6, 2019
A friend of mine gave me this book to read and it took me three attempts to finally finish reading it. For my taste, the author pushed himself, his team and his research just a little bit too much to the fore. Also, the uncritical pursuit of making teams, companies, societies etc. "better" (whatever that means) did not quite resonate with me. Methods for gathering the data needed for analysis seem - to me - to be highly intrusive into privacy. A quote that is quite symbolic for the book: "... today there isn't a single organization in the world that keeps track of both face-to-face and electronic interaction patterns. And, as we all know, what isn't measured can't be managed." Thank you, but no, thank you. Event though the "New Deal on Data" features heavily in the book and, much more so, the author's contribution to it, I'm not convinced to give up all my privacy for research.
April 6, 2014
The writing style contained a hint of sales pitch at times, but the concepts are interesting, and are founded on a number of peer reviewed journal articles. The appendices contain a lot of the mathematics, methodology, and development information, and the text and citations refer to the journal articles to delve into this in more depth, while keeping the text explanation accessible to a general audience, if still a bit jargony. All of this is predicated on "big-data" analysis and computational social science.
The author makes some interesting proposals to increase the flow of ideas and innovation, from a small group level, to within workgroups and companies, to cities, and society at large. He posits that improving idea flow will solve a wide range of society's problems. Improving idea flow includes reducing group-think where networks of people are too densely interconnected and recursive, while improving connections to isolated individuals to bring in different information. This occurs through exploration behavior--establishing connections outside your normal circles, while changing the actions and behavior to build better habits and collective intelligence in your network involves engagement. The focus is to improve the quality of collective intelligence, productivity and creativity. He describes how this can be done within organizations, and then shares a vision for data-driven cities, using sensing and agile systems to improve the flow of ideas, as well as making infrastructure more adaptive to the needs of its inhabitants.
These ideas are very interesting and compelling, but despite the assurances he would see put in place to ensure these systems are not abused, I'm still concerned. He champions private ownership of personal data, rather than corporations or the government, so that users can make informed choices about what they share to maintain their privacy. He also proposes a method to share selected information with services anonymously or selectively. Yet as this data and the tools to analyze it become more and more important, the incentives for abuse commensurately increase. He proposes using big data analytics from all of this data to make adaptive policy. But even if individuals own their personal data, and even if the process of using it is transparent, only a small number of people will be able to tell if the data is being cherry picked or how the algorithms analyzing the data works. Selfish actors could skew public or private policy in their favor behind the scenes while leading everyone to believe the outcome is in the public interest or the result of a fair practice.
Whether you are comfortable with what the author proposes or not, this is something we need to come to grips with. The author calls it "Promethean Fire" because it can warm us or burn us depending on how it's used. His vision for this phenomenon is a little utopian, and shows the promising ways it can be used, but we need to look much harder at preventing its abuse. Either way, it's happening.
The author makes some interesting proposals to increase the flow of ideas and innovation, from a small group level, to within workgroups and companies, to cities, and society at large. He posits that improving idea flow will solve a wide range of society's problems. Improving idea flow includes reducing group-think where networks of people are too densely interconnected and recursive, while improving connections to isolated individuals to bring in different information. This occurs through exploration behavior--establishing connections outside your normal circles, while changing the actions and behavior to build better habits and collective intelligence in your network involves engagement. The focus is to improve the quality of collective intelligence, productivity and creativity. He describes how this can be done within organizations, and then shares a vision for data-driven cities, using sensing and agile systems to improve the flow of ideas, as well as making infrastructure more adaptive to the needs of its inhabitants.
These ideas are very interesting and compelling, but despite the assurances he would see put in place to ensure these systems are not abused, I'm still concerned. He champions private ownership of personal data, rather than corporations or the government, so that users can make informed choices about what they share to maintain their privacy. He also proposes a method to share selected information with services anonymously or selectively. Yet as this data and the tools to analyze it become more and more important, the incentives for abuse commensurately increase. He proposes using big data analytics from all of this data to make adaptive policy. But even if individuals own their personal data, and even if the process of using it is transparent, only a small number of people will be able to tell if the data is being cherry picked or how the algorithms analyzing the data works. Selfish actors could skew public or private policy in their favor behind the scenes while leading everyone to believe the outcome is in the public interest or the result of a fair practice.
Whether you are comfortable with what the author proposes or not, this is something we need to come to grips with. The author calls it "Promethean Fire" because it can warm us or burn us depending on how it's used. His vision for this phenomenon is a little utopian, and shows the promising ways it can be used, but we need to look much harder at preventing its abuse. Either way, it's happening.
February 11, 2017
Sharp, fast, and smart book about new approach to analysis of and action in networked societies and data-rich environment. The book offers overview of a broad range of researches conducted at or led by Alex Pentland from MIT, all connected by central idea of "Social Physics". Pentland argues that central for our networked society is flow of ideas, and by regulating it we could get better results. Our Data Rich environment offers many opportunities for measuring and designing these flows. The book resonates strongly with a number of other books I've read recently on complexity and networks. For instance, what Gen. McChrystal did intuitively in Iraq (see "Team of Teams" http://amzn.to/2kdaMGh), now is measured and described in terms of ideas flow. Movement towards networked society give a new light to "The End of Average: How We Succeed in a World That Values Sameness" http://amzn.to/2kYTYSR, which argued against averaging approach for more personalized one. The "Foragers, Farmers, and Fossil Fuels: How Human Values Evolve" http://amzn.to/2kCga4d argued that energy reach Fossil Fuels society will gradually eliminate all internal barriers and segregation lines, while Pentland demonstrate how it would happen and how new dynamic tribes will form.
However, the book is not without problems. I hesitated between 5 starts for ideas and 3 starts for style. The downside of the book is its nature, this is de facto promotion pamphlet of author. It is peppered with mentioning of every company he set up and every PhD student he engaged (not necessarily successful, many references are to conference presentations, rather than books or peer reviewed papers).
http://amzn.to/2kf0O2c
However, the book is not without problems. I hesitated between 5 starts for ideas and 3 starts for style. The downside of the book is its nature, this is de facto promotion pamphlet of author. It is peppered with mentioning of every company he set up and every PhD student he engaged (not necessarily successful, many references are to conference presentations, rather than books or peer reviewed papers).
http://amzn.to/2kf0O2c
July 20, 2016
Overall I loved the book, and if Good Reads allowed for it, I'd have given it a 3.5. Maybe that sounds contradictory, let me explain.
The first half of the book is solid. How good ideas spread based social group dynamics, lateral conversations within strict, vertical hierarchies, how engagement is achieved between employees to increase productivity. If this all sounds a little dry, Pentland manages to make it lively with examples from real life applications he and his research team conducted.
The second half concerns itself with cities, and it's here I'm torn. Full disclosure: I think cities are the solutions to all of our problems. So take this with a grain of salt. Primarily, the argument in the latter half of the book is that we need to find a way to use personal data to create a public, open source meta data collection so that citizens can best assess the needs of the city.
While this is accurate, and I'd love it, we run into a few problems here. One, there's a lot of vagueness around how the data privacy is secured. Saying we need a New Deal for Data is fantastic, but achieving it is another step. Secondly, Pentland takes one too many jabs at current solutions to urban problems while having only hopes and promises to go on. As an example, he makes the assertion that congestion pricing is not only bad, it's a system that rewards the rich.
It's beyond me how congestion pricing, which taxes vehicles used within city limits, would punish the poor when they're already likely taking public transit. As for those on the fringe in the middle class, asking them to ride the train instead of paying for parking in the city may be an inconvenience, but it's hardly confiscatory. Moreover, the externalities congestion creates would be removed, which economically speaking is an additional cost savings.
It's arguments like these that appear one-sided intentionally to favor his arguments. Having said all of this, the data does display a side of an argument that is certainly worth looking into, especially if communal, open-source data is achievable in the near future.
The first half of the book is solid. How good ideas spread based social group dynamics, lateral conversations within strict, vertical hierarchies, how engagement is achieved between employees to increase productivity. If this all sounds a little dry, Pentland manages to make it lively with examples from real life applications he and his research team conducted.
The second half concerns itself with cities, and it's here I'm torn. Full disclosure: I think cities are the solutions to all of our problems. So take this with a grain of salt. Primarily, the argument in the latter half of the book is that we need to find a way to use personal data to create a public, open source meta data collection so that citizens can best assess the needs of the city.
While this is accurate, and I'd love it, we run into a few problems here. One, there's a lot of vagueness around how the data privacy is secured. Saying we need a New Deal for Data is fantastic, but achieving it is another step. Secondly, Pentland takes one too many jabs at current solutions to urban problems while having only hopes and promises to go on. As an example, he makes the assertion that congestion pricing is not only bad, it's a system that rewards the rich.
It's beyond me how congestion pricing, which taxes vehicles used within city limits, would punish the poor when they're already likely taking public transit. As for those on the fringe in the middle class, asking them to ride the train instead of paying for parking in the city may be an inconvenience, but it's hardly confiscatory. Moreover, the externalities congestion creates would be removed, which economically speaking is an additional cost savings.
It's arguments like these that appear one-sided intentionally to favor his arguments. Having said all of this, the data does display a side of an argument that is certainly worth looking into, especially if communal, open-source data is achievable in the near future.
July 18, 2014
"The most productive people are constantly developing and testing a new story, adding newly discovered ideas to the story and then trying it out on everyone they meet. Like sculpting raw clay into a beautiful statue, over time their story becomes more and more compelling. Finally they decide that it is time to act on it, to bring it into the light and test it against reality. To these people, the practice of harvesting, winnowing, and sculpting ideas feels like play. In fact, some of them call it 'serious play.'" (26-7)
"On average, it turned out that the social network incentive scheme worked almost four times more efficiently than a traditional individual-incentive market approach. For the buddies that had the most interactions with their assigned target, the social network incentive worked almost eight times better than the standard market approach." (on the FunFit experiment, 68)
"This social network incentive [performed by a Swiss utility] caused electricity consumption to drop by 17 percent, twice the best result seen in earlier conservation campaigns and more than four times more effective than the typical energy reduction campaign." (72)
"These bursts of exploration -- shopping trips, days off that are spent wandering around the city, weekend getaways -- seem to be important in growing the local ecology of cities. If we looked at cities with greater than average rates of exploration in the credit card data, we found that in subsequent years they had a higher GDP, a larger population, and a greater variety of stores and restaurants. It makes sense that more exploration, which results in a greater number of interactions between current norms and new ideas, would be a driver of innovative behavior." (162-3)
"[O]n-the-fly reduction of the dimensionality and scope of the data so that only the minimum needed for the specific problem is shared makes things safer. Such a mechanism also allows users to safely grant and revoke data access, to share data anonymously without needing a trusted third party, and to monitor and audit data uses." (229)
"On average, it turned out that the social network incentive scheme worked almost four times more efficiently than a traditional individual-incentive market approach. For the buddies that had the most interactions with their assigned target, the social network incentive worked almost eight times better than the standard market approach." (on the FunFit experiment, 68)
"This social network incentive [performed by a Swiss utility] caused electricity consumption to drop by 17 percent, twice the best result seen in earlier conservation campaigns and more than four times more effective than the typical energy reduction campaign." (72)
"These bursts of exploration -- shopping trips, days off that are spent wandering around the city, weekend getaways -- seem to be important in growing the local ecology of cities. If we looked at cities with greater than average rates of exploration in the credit card data, we found that in subsequent years they had a higher GDP, a larger population, and a greater variety of stores and restaurants. It makes sense that more exploration, which results in a greater number of interactions between current norms and new ideas, would be a driver of innovative behavior." (162-3)
"[O]n-the-fly reduction of the dimensionality and scope of the data so that only the minimum needed for the specific problem is shared makes things safer. Such a mechanism also allows users to safely grant and revoke data access, to share data anonymously without needing a trusted third party, and to monitor and audit data uses." (229)
February 18, 2017
A good place to start for anyone who would like to know more about Sandy's amazing work in understanding social connections. The writing is a bit dry so do treat it as a longish review paper and not a pop science novel.
April 15, 2018
在網上有一篇很完整的書評 在此分享 有標出處和作者
https://book.douban.com/review/7804309/
作者:
倪考梦 2016-03-08 13:49:20
个体智慧与群体智慧
1、本文系笔者在温州市决策咨询办公室内部学习会上的发言提纲，篇幅较长，结构粗糙，且有删节，敬请大家谅解。
2、本文也是笔者《智慧社会》一书的读后感，根据湛庐『庐客汇』网友建议公布分享，欢迎大家多提出意见与建议，帮助我完善关于个体智慧和群体智慧的思考。
－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－
作为智库，我希望大家都拥有强大的个人智慧，同时，作为整体，我们拥有超过每个个体的群体智慧。这样，我们才能为思想市场提供更高水平的思想产品。为此，我们需要去了解他们的工作机制。
一、神经网络与社会网络是同构的
根据复杂理论（complexity theory）的观点，这些发明与创新的模式是分形（fractal）的，即当我们放大或缩小比例进行观察时，这些模式会以一种可辨认的形态呈现出来，不管是分子、神经元、像素，还是人行道都会再现同一种模式。从逻辑上说，神经网络与社会网络是同构的，都是复杂系统。《智慧社会》提出的从想法流的角度来看社会网络，给我很大的启发。他是研究神经科学的计算机科学家，他对社会网络的研究中隐含了他来自神经网络研究的各种类比。
对于个体智慧而言，神经网络是想法机器，神经网络强调是意志作为连接者，对外探索，对内联想，形成个体智慧，呈现为意识中的思维。
对于群体智慧而言，社会网络也是想法机器，社会网络中有魅力型连接者组织群体成员对外探索，对内参与，形成群体智慧。
二、探索和参与带来想法的流通
1、何谓探索和参与
探索即挖点，或者说个人的学习，群体的学习。参与即连线，或者说个体的思考，群体的交流。古话说，“终日而思矣，不如须臾之所学”，“温故而知新”，“学而不思则罔，思而不学则殆”，都是这个逻辑。
2、探索和参与的理想比重
《智慧社会》一书中提出一个量化的观点，就是探索和参与应该是9：1。我们考虑参考这个比例来做研究。换句话说，以后我们要更多地探索。
3、探索和参与的分时
我个人的策略有二：一是在清晨时间做好个人学习。晨型阶段是属于我自己的，我要做的是独立探索，自由联想，优化神经网络，提高个人能力。二是在其他时间做好群体学习。其他时间则是属于社会的，我要做的是集体探索，集体参与，优化社会网络，提高群体智能。这个过程中我还要扮演好魅力型连接者的角色，要为我的团队良性探索和参与创造好的条件。
三、探索和参与的基本原则
1、简单问题与复杂问题
为了方便思考，我引入《预知社会》《完美的群体》《群体的智慧》中的观点：一是个人的不可预测性，会在集体行为中被降低。人数越多，个人的意愿就越会深埋在普通事实的系列之下。二是无论我们是否采用平均数，是否接纳大多数人的意见，或是否找出知识最渊博的人并听取他们的指导，总是有一些人要牺牲自己的意见以使整个群体能够从群体智慧中获益。三是针对有标准答案的『简单问题』，参与的人多了以后，正确答案固化下来，噪音则相互抵消，所以多数人的共识更接近于正确答案。这里有个原则，“填数值”请用“均值”，“单选题”请用“多数”。比如说，问大家月球到地球距离，用大家答案的平均值会更接近于正确答案。又比如说，问大家二选一的简单问题，用多数人答案更可能答对。四是对于没有标准答案的『复杂问题』，专家智慧强于群体智慧。
我们在讨论探索和参与之前，从问题导向出发的话，首先要定义清楚问题。或者先定义清楚目标，然后找出问题。接着，对问题进行分析，看它到底是属于哪一种。不同的问题有不同的智能解决方案。
根据之前那些书的观点，我们可以假设：简单问题要下问（问大众）。复杂问题要上问（问专家）。
不过不管是哪一类问题，问都比不问好，因为问就是一个探索的过程。只要是我们不知道答案的问题，问了总会有一些线索帮助我们更快捷的找到答案。当然，这里也存在只问一个人，然后问错人的情况。
2、快搜索与慢搜索
快搜索其实就是只问大众。慢搜索则是问多个专家。如果结合上述两类问题的思考，则出现四种情况：一是简单问题×快搜索，直接用百度（代表大众共识），可以给出很好的答案。如百度很清楚1+1=2。二是简单问题×慢搜索，最后可能效果不好，因为专家会把简单的问题复杂化。如你问专家1+1=?，或许会有很多答案。三是复杂问题×快搜索，直接用百度或者谷歌，会得到大众的偏见，或者专家的偏见，相对而言，专家的会靠谱一点，但是我们如果没有能力识别专家的话，就变成碰运气了。四是复杂问题×慢搜索，先通过专业渠道（线上专业搜索引擎或者线下智库机构咨询专家），相对而言，获得的信息质量会高一些。我的观点是：遇到问题时，先分类。然后简单问题用快搜索，复杂问题用慢搜索。
（１）简单问题×快探索。在只有手机或者时间有限的情况下，只能快速检索一些与问题相关的信息。我把这种状态成为快探索。生活中大部分情况下都是快探索，人脑自动读取经验来操作，就是一种快探索。这里只能是满意策略。为了提高快探索的质量，就必须事先设计好方法。由于资源有限，所以一般来说，快探索的第一入口肯定是搜索引擎，但是为了提高效率，必须使用正确的搜索引擎。对于简单问题，可以直接搜索百度，搜不到，再升级为Google。由于问题本身简单，所以基本都会搞定。
（２）复杂问题×慢搜索。复杂问题往往没有简单的答案、精准的答案，甚至可能就没有答案。大部分我要研究的问题，都属于这类。我想到大概有三步：一是用好大脑。我必须先用纸张画思维导图，在探索自己大脑信息的过程中，结合Evernote里保存的历史笔记，梳理逻辑。二是用好网络。专用搜索（如Flickr、知乎、轻单等）到综合搜索（Google、百度等），逐步升级。——为了让复杂问题能有好的答案，我平日里就必须在探索过程中，不断积累优质信息，尤其是一手信息。这些信息如果是直接搜索综合引擎，肯定是很难梳理出来，但如果我自己平日里积累了，那么就会很快找到他们。三是用好专家。如果自己想不出来，也搜索不到，就打电话请教专家，或者上网请教专家。这是给他人投资我的机会，也会强化彼此之间的社会互动。
总之，在找到答案之后，再开始解决问题。
3、快参与与慢参与
这个问题我之前没思考过。如果让我思考，我想大概就是思想酝酿的充分程度了。
对个体而言，参与就是一个神经网络消化新信息的过程。如果是快思考，则必然会偏向于关联思维，就是在脑子里自动加工，然后生成一个印象。如果是慢思考，则会生成更为复杂的结果。快参与是即使决断，慢参与则是有意志参与的过程，这个不难理解。
对于群体而言，参与就很复杂了。在方式不变的情况下，时间往往和参与程度成正比。快参与或许就是我简单把我了解到的情况跟大家说一下，然后听下大家的反馈。时间仓促自然参与程度不深，大家未必能够消化这些信息，并给出高质量的反馈，所以快参与未必能激发超越个体的群体智慧。当然，有比没有好，这个方式对于群体中的其他个体来说，都是一个学习的过程，会让他们受益。我现在跟单位里的同事们开短会，讲解一些知识，都属于快参与。这个过程好比吃快餐，可以填饱肚子但体验不佳。
慢参与理论上说会好得多，就是大家投入时间去慢慢消化这些信息，充分地互动，比如说专门跑到一个地方去，专心在这里相互交流。对于做课题来说，就是围绕某个专题进行深入的座谈。我们大部分的座谈会照理说都要按照这个标准去做，只是大家还做不到罢了。在现实操作中，确保每个人都和每个人交谈很困难，但是推动每个人都发言是完全可以做到的。我现在能想到的就是每次会议都让每个人谈一个主题，然后其他每个人都对这个主题给出意见和建议，通过制度化的方式来增进参与。如此一来，才像是思想的大餐。
此外，闲聊也可以作为一种参与方式。我要询问身边的同事他们的情况，倾听他们的收获，再给予一定的回馈。在这个过程中，自己收获一些信息，同时也强化一些认知，提炼一些知识，并帮助他人。
四、群体智慧：社会网络的优化思路
从社会网络角度说，我要考虑的是大家群体智慧如何发挥。……将来如果有条件的话，或许我会去尝试使用《智慧社会》里提到的社会计量标牌和Funf智能应用。
我们这些人的特点就是，理论水平不如高校学者，实践经验不如政府官员；但反过来说，我们的实践经验多于高校专家，理论水平高于政府官员。对大家来说，如果能挖好点，连好线，成为蜜蜂，就可以兼顾高校学者的理论和政府官员的经验，具备优势。
1、外部探索
就外部探索而言，我想主要是要激励大家更多地与外部交流，多调研，多参会，多电话，多看书。差不多就是把我个人的内容转移到他们身上。作为魅力型连接者，或者说主持人、中场组织者、蜜蜂，我可以帮他们对接其他人，或者组织调研，或者提供教学，或者交代任务，让他们在工作中不得不接触各种人，收获各种信息。就今年而言，我认为调研是最好的形式，看书也是一种。记忆和文字是跨时空连接人和想法，读书是探索另一个时空的人和想法。调研也是。
2、内部参与
《伟大创意的诞生》一书认为，相比于将创意保护起来，连接创意会让我们更有收获。
（１）我们的学习会是最好的面对面参与的方式，只是在会议规则方面还需要进一步设计。现在的会议比较死板，不利于他们彼此之间走动，改变位置，相互交流。所以下一步开学习会，最好是在舒服的地方，比如说咖啡馆，大家轻松一点。自由的座位，自由的交流，这���关键。沙龙也是如此，以后要把自由讨论环节加上去，压缩前面的发言时间。由于社会影响对他人有推动，言传不如身教，所以主持人要努力推动大家之间彼此交流，并主动发言。
（２）团队规模很重要。尤其是群聊的情况下，如果人多了，基本上就残废了。……以后团队大了，真的有必要把它拆开。
（３）课题研究也是如此，团队人要少要精，分开来探索信息，然后合起来交流信息。……
（４）结合《伟大创意的诞生》，面对面的参与需要我们重新设计环境，让人们的想法可以在很多场所交流和溢出。……
思考下各类工作的重新设计。一是会议。我想流程就要改变，要设法让大家可以相互交流，自由讨论是比较好的方式，我的角色就是主持人。为此需要每次会议让大家准备好议题，给予足够的时间来交流。这里组织好���群体智慧就高，组织不好群体智慧就低。二是文件。原理也是一样，如果是增加一个讨论的环节，那么可能会获得一些有用的信息。
五、个体智慧：神经网络的优化思路
个人层面，我认为有意识的意志可以发挥一个重要的作用，就是事先去设计好探索和联想的策略。我称之为探索+和联想+，或者说探索家计划和梦想家计划。
这里先讨论一下意志的能动性。根据先前的思考，慢思考要发现问题，制订计划，有意识的意志可以明确目标，限制边界，引入信息工具（如evernote化的外脑），有限试错（改变要素，改变结构），杠铃策略，满意策略（见好就收），最后不断迭代。意志可以调控思维形成过程中的每个要素及其关系。比如说改变外部信息，改变内部信息，改变他们之间的组合关系（通过否决当前方案）。快思考的任务是不断试错，随机游走，找到好的解决方案。这是一个漫长积累，一朝顿悟的过程。
1、探索家计划（发现新组件）
探索家计划包含之前我思考过的各种探索方式，包括阅读、实践、调研、内省等。为了提高效率，必须发挥意志的作用。
（１）明确目标。我探索的目的是为了达成什么效果。
（２）限制边界。缩小探索的范围，提高效能。
（３）引入工具。将evernote等工具引进来，提高效率，建立『外脑』。探索要做好成果的整理和记录。这里我想到还是用evernote最好。为了让更多的信息能够在联想阶段被关联起来，我认为有必要把我的日记里的断想，读书笔记，调研笔记，工作笔记等，还有以前发过的文章，所有的文字，都存入evernote，由此形成一个真正意义上的外脑，或者说个人的大数据库。这个逻辑还可以引申到我们智库的知识管理。
（４）杠铃策略，有限试错。
一是选人策略。人是重要的社会网络节点，找到人往往就找到了优秀的信息源。我想可以把人分为三类：第一等是聪明人，就是《智慧社会》里说的，能预测他人行为，然后做出不同决断的人，这些人表现得很独立，他们的意见非常值得参考。……第二等是实干家，就是有思想，又掌握资源，做着某个具体项目的人，包括各种精英，比如说企业家，还有一些大专家。对这类人，我们的策略就是找出每个网络里最重要的节点就好了。比如说，……此外，《决断力》一书认为，我们要“听一线专家对形势的分析，但不要相信他们对未来的判断”。这个观点可以参考。第三等是空谈者，这里人很多了，稍微有点想法的，都可以归入此类，对于这类人，只能是弱关系了，但是如果有一个规模的话，将来可以发挥重要的信息提供者的作用。所以要多多益善。从这个角度说，在任何场合和任何陌生人建立关联，还是有价值的。当然，前提是没有风险的情况下。第一类人我们要建立起强关系，第二、三类则是大规模铺开。要讨论问题，就找第一类人交流。……在探索的过程中，请教人或者帮助人，都属于基本的互动方式，不在于内容，而在于行动本身。只要有交流，就会强化连接，这和神经网络应该是相似的。我的策略可以简化为，和几个聪明人建立强关系，……经常讨论各种问题，有意识地强化这个网络；与其他人维持弱关系，全部纳入微信朋友圈，根据需求来请教人和帮助人，无意识地强化这个网络。
二是选书、选讲座、选调研活动的策略。我要参考两个参数，一是内容多样性，是否新。二是专家独立性，是否强。这里形成四种情况。第一等是内容新，专家强，必选。第二等是内容旧，专家强，也可以考虑。第三等是内容新，专家弱，也可以考虑。第四等是内容旧、专家弱，舍弃。有了这个标准，以后买书时就好选择了，尤其是在不同的内容之间，就有一个明确的标准做出快速选择了。
（５）满意策略，见好就收。然后根据实际，还要不断迭代，来更新信息。包容错误。努力把错误变成创见。容错率的背后是保持冗余来应对错误可能带来的风险冲击。
2、梦想家计划（创造新组合）
梦想家计划也要发挥意志的作用，也需要明确目标，限制边界，引入工具，有限试错，见好就收。梦想家计划的目的是给原本不相干的东西建立关联，获得连接者的红利，这是一个收割结构洞的过程。这里我想到一些可以马上去做的事情。
（１）处理好主线任务和支线任务的关系。我可以同时做多个任务来获得多样性的信息，然后加以联想，但是不要过多。任务切换时，应以主线任务为主，快速记录与支线任务有关的碎片想法（灵感），然后回到主线任务继续前进，直至注意力弱化了，再切换任务。这个在阅读书籍的时候特别明显，因为看着看着很容易想到一些东西，这时候快速用本子记录下断想，然后回归书本，相对是好的策略。这样做的目的，是为了降低意志力的消耗。
（２）多做笔记。用大本子来扩展内存，本子的纸张大了，能写的范围大了，思维会更开阔。我可以在写完后把它数字化，然后碎掉。
（３）以否定促创新。在日记里测试思辨模式，设计一个提问者，批判态度，一个回答者，创新态度，然后平等交流，在这个过程中不断批判现有的想法，迫使自己思维提升。提问者可以是苏格拉底，回答者可以是富兰克林。这个过程其实是将自己的思维外部化，推动不同神经网络之间的互动，这个过程会让一个想法变得更加深刻。
（写于2015年4月16日，转自倪考梦的微信公众号）
https://book.douban.com/review/7804309/
作者:
倪考梦 2016-03-08 13:49:20
个体智慧与群体智慧
1、本文系笔者在温州市决策咨询办公室内部学习会上的发言提纲，篇幅较长，结构粗糙，且有删节，敬请大家谅解。
2、本文也是笔者《智慧社会》一书的读后感，根据湛庐『庐客汇』网友建议公布分享，欢迎大家多提出意见与建议，帮助我完善关于个体智慧和群体智慧的思考。
－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－
作为智库，我希望大家都拥有强大的个人智慧，同时，作为整体，我们拥有超过每个个体的群体智慧。这样，我们才能为思想市场提供更高水平的思想产品。为此，我们需要去了解他们的工作机制。
一、神经网络与社会网络是同构的
根据复杂理论（complexity theory）的观点，这些发明与创新的模式是分形（fractal）的，即当我们放大或缩小比例进行观察时，这些模式会以一种可辨认的形态呈现出来，不管是分子、神经元、像素，还是人行道都会再现同一种模式。从逻辑上说，神经网络与社会网络是同构的，都是复杂系统。《智慧社会》提出的从想法流的角度来看社会网络，给我很大的启发。他是研究神经科学的计算机科学家，他对社会网络的研究中隐含了他来自神经网络研究的各种类比。
对于个体智慧而言，神经网络是想法机器，神经网络强调是意志作为连接者，对外探索，对内联想，形成个体智慧，呈现为意识中的思维。
对于群体智慧而言，社会网络也是想法机器，社会网络中有魅力型连接者组织群体成员对外探索，对内参与，形成群体智慧。
二、探索和参与带来想法的流通
1、何谓探索和参与
探索即挖点，或者说个人的学习，群体的学习。参与即连线，或者说个体的思考，群体的交流。古话说，“终日而思矣，不如须臾之所学”，“温故而知新”，“学而不思则罔，思而不学则殆”，都是这个逻辑。
2、探索和参与的理想比重
《智慧社会》一书中提出一个量化的观点，就是探索和参与应该是9：1。我们考虑参考这个比例来做研究。换句话说，以后我们要更多地探索。
3、探索和参与的分时
我个人的策略有二：一是在清晨时间做好个人学习。晨型阶段是属于我自己的，我要做的是独立探索，自由联想，优化神经网络，提高个人能力。二是在其他时间做好群体学习。其他时间则是属于社会的，我要做的是集体探索，集体参与，优化社会网络，提高群体智能。这个过程中我还要扮演好魅力型连接者的角色，要为我的团队良性探索和参与创造好的条件。
三、探索和参与的基本原则
1、简单问题与复杂问题
为了方便思考，我引入《预知社会》《完美的群体》《群体的智慧》中的观点：一是个人的不可预测性，会在集体行为中被降低。人数越多，个人的意愿就越会深埋在普通事实的系列之下。二是无论我们是否采用平均数，是否接纳大多数人的意见，或是否找出知识最渊博的人并听取他们的指导，总是有一些人要牺牲自己的意见以使整个群体能够从群体智慧中获益。三是针对有标准答案的『简单问题』，参与的人多了以后，正确答案固化下来，噪音则相互抵消，所以多数人的共识更接近于正确答案。这里有个原则，“填数值”请用“均值”，“单选题”请用“多数”。比如说，问大家月球到地球距离，用大家答案的平均值会更接近于正确答案。又比如说，问大家二选一的简单问题，用多数人答案更可能答对。四是对于没有标准答案的『复杂问题』，专家智慧强于群体智慧。
我们在讨论探索和参与之前，从问题导向出发的话，首先要定义清楚问题。或者先定义清楚目标，然后找出问题。接着，对问题进行分析，看它到底是属于哪一种。不同的问题有不同的智能解决方案。
根据之前那些书的观点，我们可以假设：简单问题要下问（问大众）。复杂问题要上问（问专家）。
不过不管是哪一类问题，问都比不问好，因为问就是一个探索的过程。只要是我们不知道答案的问题，问了总会有一些线索帮助我们更快捷的找到答案。当然，这里也存在只问一个人，然后问错人的情况。
2、快搜索与慢搜索
快搜索其实就是只问大众。慢搜索则是问多个专家。如果结合上述两类问题的思考，则出现四种情况：一是简单问题×快搜索，直接用百度（代表大众共识），可以给出很好的答案。如百度很清楚1+1=2。二是简单问题×慢搜索，最后可能效果不好，因为专家会把简单的问题复杂化。如你问专家1+1=?，或许会有很多答案。三是复杂问题×快搜索，直接用百度或者谷歌，会得到大众的偏见，或者专家的偏见，相对而言，专家的会靠谱一点，但是我们如果没有能力识别专家的话，就变成碰运气了。四是复杂问题×慢搜索，先通过专业渠道（线上专业搜索引擎或者线下智库机构咨询专家），相对而言，获得的信息质量会高一些。我的观点是：遇到问题时，先分类。然后简单问题用快搜索，复杂问题用慢搜索。
（１）简单问题×快探索。在只有手机或者时间有限的情况下，只能快速检索一些与问题相关的信息。我把这种状态成为快探索。生活中大部分情况下都是快探索，人脑自动读取经验来操作，就是一种快探索。这里只能是满意策略。为了提高快探索的质量，就必须事先设计好方法。由于资源有限，所以一般来说，快探索的第一入口肯定是搜索引擎，但是为了提高效率，必须使用正确的搜索引擎。对于简单问题，可以直接搜索百度，搜不到，再升级为Google。由于问题本身简单，所以基本都会搞定。
（２）复杂问题×慢搜索。复杂问题往往没有简单的答案、精准的答案，甚至可能就没有答案。大部分我要研究的问题，都属于这类。我想到大概有三步：一是用好大脑。我必须先用纸张画思维导图，在探索自己大脑信息的过程中，结合Evernote里保存的历史笔记，梳理逻辑。二是用好网络。专用搜索（如Flickr、知乎、轻单等）到综合搜索（Google、百度等），逐步升级。——为了让复杂问题能有好的答案，我平日里就必须在探索过程中，不断积累优质信息，尤其是一手信息。这些信息如果是直接搜索综合引擎，肯定是很难梳理出来，但如果我自己平日里积累了，那么就会很快找到他们。三是用好专家。如果自己想不出来，也搜索不到，就打电话请教专家，或者上网请教专家。这是给他人投资我的机会，也会强化彼此之间的社会互动。
总之，在找到答案之后，再开始解决问题。
3、快参与与慢参与
这个问题我之前没思考过。如果让我思考，我想大概就是思想酝酿的充分程度了。
对个体而言，参与就是一个神经网络消化新信息的过程。如果是快思考，则必然会偏向于关联思维，就是在脑子里自动加工，然后生成一个印象。如果是慢思考，则会生成更为复杂的结果。快参与是即使决断，慢参与则是有意志参与的过程，这个不难理解。
对于群体而言，参与就很复杂了。在方式不变的情况下，时间往往和参与程度成正比。快参与或许就是我简单把我了解到的情况跟大家说一下，然后听下大家的反馈。时间仓促自然参与程度不深，大家未必能够消化这些信息，并给出高质量的反馈，所以快参与未必能激发超越个体的群体智慧。当然，有比没有好，这个方式对于群体中的其他个体来说，都是一个学习的过程，会让他们受益。我现在跟单位里的同事们开短会，讲解一些知识，都属于快参与。这个过程好比吃快餐，可以填饱肚子但体验不佳。
慢参与理论上说会好得多，就是大家投入时间去慢慢消化这些信息，充分地互动，比如说专门跑到一个地方去，专心在这里相互交流。对于做课题来说，就是围绕某个专题进行深入的座谈。我们大部分的座谈会照理说都要按照这个标准去做，只是大家还做不到罢了。在现实操作中，确保每个人都和每个人交谈很困难，但是推动每个人都发言是完全可以做到的。我现在能想到的就是每次会议都让每个人谈一个主题，然后其他每个人都对这个主题给出意见和建议，通过制度化的方式来增进参与。如此一来，才像是思想的大餐。
此外，闲聊也可以作为一种参与方式。我要询问身边的同事他们的情况，倾听他们的收获，再给予一定的回馈。在这个过程中，自己收获一些信息，同时也强化一些认知，提炼一些知识，并帮助他人。
四、群体智慧：社会网络的优化思路
从社会网络角度说，我要考虑的是大家群体智慧如何发挥。……将来如果有条件的话，或许我会去尝试使用《智慧社会》里提到的社会计量标牌和Funf智能应用。
我们这些人的特点就是，理论水平不如高校学者，实践经验不如政府官员；但反过来说，我们的实践经验多于高校专家，理论水平高于政府官员。对大家来说，如果能挖好点，连好线，成为蜜蜂，就可以兼顾高校学者的理论和政府官员的经验，具备优势。
1、外部探索
就外部探索而言，我想主要是要激励大家更多地与外部交流，多调研，多参会，多电话，多看书。差不多就是把我个人的内容转移到他们身上。作为魅力型连接者，或者说主持人、中场组织者、蜜蜂，我可以帮他们对接其他人，或者组织调研，或者提供教学，或者交代任务，让他们在工作中不得不接触各种人，收获各种信息。就今年而言，我认为调研是最好的形式，看书也是一种。记忆和文字是跨时空连接人和想法，读书是探索另一个时空的人和想法。调研也是。
2、内部参与
《伟大创意的诞生》一书认为，相比于将创意保护起来，连接创意会让我们更有收获。
（１）我们的学习会是最好的面对面参与的方式，只是在会议规则方面还需要进一步设计。现在的会议比较死板，不利于他们彼此之间走动，改变位置，相互交流。所以下一步开学习会，最好是在舒服的地方，比如说咖啡馆，大家轻松一点。自由的座位，自由的交流，这���关键。沙龙也是如此，以后要把自由讨论环节加上去，压缩前面的发言时间。由于社会影响对他人有推动，言传不如身教，所以主持人要努力推动大家之间彼此交流，并主动发言。
（２）团队规模很重要。尤其是群聊的情况下，如果人多了，基本上就残废了。……以后团队大了，真的有必要把它拆开。
（３）课题研究也是如此，团队人要少要精，分开来探索信息，然后合起来交流信息。……
（４）结合《伟大创意的诞生》，面对面的参与需要我们重新设计环境，让人们的想法可以在很多场所交流和溢出。……
思考下各类工作的重新设计。一是会议。我想流程就要改变，要设法让大家可以相互交流，自由讨论是比较好的方式，我的角色就是主持人。为此需要每次会议让大家准备好议题，给予足够的时间来交流。这里组织好���群体智慧就高，组织不好群体智慧就低。二是文件。原理也是一样，如果是增加一个讨论的环节，那么可能会获得一些有用的信息。
五、个体智慧：神经网络的优化思路
个人层面，我认为有意识的意志可以发挥一个重要的作用，就是事先去设计好探索和联想的策略。我称之为探索+和联想+，或者说探索家计划和梦想家计划。
这里先讨论一下意志的能动性。根据先前的思考，慢思考要发现问题，制订计划，有意识的意志可以明确目标，限制边界，引入信息工具（如evernote化的外脑），有限试错（改变要素，改变结构），杠铃策略，满意策略（见好就收），最后不断迭代。意志可以调控思维形成过程中的每个要素及其关系。比如说改变外部信息，改变内部信息，改变他们之间的组合关系（通过否决当前方案）。快思考的任务是不断试错，随机游走，找到好的解决方案。这是一个漫长积累，一朝顿悟的过程。
1、探索家计划（发现新组件）
探索家计划包含之前我思考过的各种探索方式，包括阅读、实践、调研、内省等。为了提高效率，必须发挥意志的作用。
（１）明确目标。我探索的目的是为了达成什么效果。
（２）限制边界。缩小探索的范围，提高效能。
（３）引入工具。将evernote等工具引进来，提高效率，建立『外脑』。探索要做好成果的整理和记录。这里我想到还是用evernote最好。为了让更多的信息能够在联想阶段被关联起来，我认为有必要把我的日记里的断想，读书笔记，调研笔记，工作笔记等，还有以前发过的文章，所有的文字，都存入evernote，由此形成一个真正意义上的外脑，或者说个人的大数据库。这个逻辑还可以引申到我们智库的知识管理。
（４）杠铃策略，有限试错。
一是选人策略。人是重要的社会网络节点，找到人往往就找到了优秀的信息源。我想可以把人分为三类：第一等是聪明人，就是《智慧社会》里说的，能预测他人行为，然后做出不同决断的人，这些人表现得很独立，他们的意见非常值得参考。……第二等是实干家，就是有思想，又掌握资源，做着某个具体项目的人，包括各种精英，比如说企业家，还有一些大专家。对这类人，我们的策略就是找出每个网络里最重要的节点就好了。比如说，……此外，《决断力》一书认为，我们要“听一线专家对形势的分析，但不要相信他们对未来的判断”。这个观点可以参考。第三等是空谈者，这里人很多了，稍微有点想法的，都可以归入此类，对于这类人，只能是弱关系了，但是如果有一个规模的话，将来可以发挥重要的信息提供者的作用。所以要多多益善。从这个角度说，在任何场合和任何陌生人建立关联，还是有价值的。当然，前提是没有风险的情况下。第一类人我们要建立起强关系，第二、三类则是大规模铺开。要讨论问题，就找第一类人交流。……在探索的过程中，请教人或者帮助人，都属于基本的互动方式，不在于内容，而在于行动本身。只要有交流，就会强化连接，这和神经网络应该是相似的。我的策略可以简化为，和几个聪明人建立强关系，……经常讨论各种问题，有意识地强化这个网络；与其他人维持弱关系，全部纳入微信朋友圈，根据需求来请教人和帮助人，无意识地强化这个网络。
二是选书、选讲座、选调研活动的策略。我要参考两个参数，一是内容多样性，是否新。二是专家独立性，是否强。这里形成四种情况。第一等是内容新，专家强，必选。第二等是内容旧，专家强，也可以考虑。第三等是内容新，专家弱，也可以考虑。第四等是内容旧、专家弱，舍弃。有了这个标准，以后买书时就好选择了，尤其是在不同的内容之间，就有一个明确的标准做出快速选择了。
（５）满意策略，见好就收。然后根据实际，还要不断迭代，来更新信息。包容错误。努力把错误变成创见。容错率的背后是保持冗余来应对错误可能带来的风险冲击。
2、梦想家计划（创造新组合）
梦想家计划也要发挥意志的作用，也需要明确目标，限制边界，引入工具，有限试错，见好就收。梦想家计划的目的是给原本不相干的东西建立关联，获得连接者的红利，这是一个收割结构洞的过程。这里我想到一些可以马上去做的事情。
（１）处理好主线任务和支线任务的关系。我可以同时做多个任务来获得多样性的信息，然后加以联想，但是不要过多。任务切换时，应以主线任务为主，快速记录与支线任务有关的碎片想法（灵感），然后回到主线任务继续前进，直至注意力弱化了，再切换任务。这个在阅读书籍的时候特别明显，因为看着看着很容易想到一些东西，这时候快速用本子记录下断想，然后回归书本，相对是好的策略。这样做的目的，是为了降低意志力的消耗。
（２）多做笔记。用大本子来扩展内存，本子的纸张大了，能写的范围大了，思维会更开阔。我可以在写完后把它数字化，然后碎掉。
（３）以否定促创新。在日记里测试思辨模式，设计一个提问者，批判态度，一个回答者，创新态度，然后平等交流，在这个过程中不断批判现有的想法，迫使自己思维提升。提问者可以是苏格拉底，回答者可以是富兰克林。这个过程其实是将自己的思维外部化，推动不同神经网络之间的互动，这个过程会让一个想法变得更加深刻。
（写于2015年4月16日，转自倪考梦的微信公众号）
December 21, 2013
Very interesting book. Advances the seemingly obvious claim that the increased flow of ideas between and among human agents (whether, e.g., in certain businesses, closed markets, cities, or societies) conduces to the more efficient functioning of the systems they're embedded in. Makes the further claim that idea flow increases as a function of a small number of variables, such as the relative level of engagement of agents with each other, and their relative level of exposure to the innovative ideas of other agents (especially those outside their networks). What's most interesting about the book are the empirical claims it makes, and the tests the author, his students, and his colleagues have devised to generate them. Social Physics gives credence to the idea that big data analysis truly could/might help build a better world. Social Physics is not a blueprint for doing so, nor the last word, but it's a hugely suggestive call to arms for specialists and laymen alike to pursue further studies in this area. My only criticism: there was a fair amount of redundancy in the telling. Even so, Social Physics strikes me as an essential read. Pentland and friends are clearly on to something...
February 17, 2014
The key take-aways about the ideal conditions for effective idea flow dovetail nicely with other notable recent works on group dynamics and innovation (such as The Rainforest). Organizations that allow their members to be prolific explorers of diverse ideas and provide them with rich opportunities to exchange and expose each other to those ideas will thrive. Not a surprising finding there, but the research-based insights into how face-to-face and virtual peer networks may be incentivized to accomplish these objectives are an important piece of puzzle. The mathematical models may strike some readers as the epitome of Henri Bergson's definition of comedy as "something mechanical encrusted upon the living," but the real-world experimental research bears it out. However, the biggest deal here is how eerily close the possibilities and policy implications of Social Physics are to the philosophical underpinnings of the true believers in Dave Eggers' The Circle. Here lie the tools to take us from here to there, whether that excites you or scares the hell out of you.
April 28, 2014
Social Physics is an excellent book on how to create an innovative culture in your organization. The big take-a-ways:
Creative people do the following steps
1. Absorb ideas from diverse sources
2. "Test these ideas" on lots of smart people and seeing which ideas pass their BS test.
3. Taking these new ideas that come out of the tests and testing them on more diverse people
Ideas can only "flow" by people communicating them with others.
Face to face meetings are the strongest way to create a connection between two people AND ensure that they will share ideas.
People need diverse sets of people to share their ideas with.
In a meeting when ideas are being shared, each person must be allowed to speak. If one person hogs the time by talking too much, ideas will not be shared.
Conversation turn taking is critical.
The more women who attend a meeting the more the meeting will tend to generate creative ideas by ensuring that everyone gets a voice in the discussion.
Star performers build networks and make people feel part of a team.
Creative people do the following steps
1. Absorb ideas from diverse sources
2. "Test these ideas" on lots of smart people and seeing which ideas pass their BS test.
3. Taking these new ideas that come out of the tests and testing them on more diverse people
Ideas can only "flow" by people communicating them with others.
Face to face meetings are the strongest way to create a connection between two people AND ensure that they will share ideas.
People need diverse sets of people to share their ideas with.
In a meeting when ideas are being shared, each person must be allowed to speak. If one person hogs the time by talking too much, ideas will not be shared.
Conversation turn taking is critical.
The more women who attend a meeting the more the meeting will tend to generate creative ideas by ensuring that everyone gets a voice in the discussion.
Star performers build networks and make people feel part of a team.
June 15, 2014
This book is different from most social theory books as instead of simply promoting ways of communication and how make society work better cooperatively, it uses big data to prove if methods work or not. I wouldn't say it's ground breaking, but offers a unique perspective that hasn't been offered before.
My complaint with the book is that near the end, it felt like a doctorate thesis about looser privacy regulations so he can get better access to individuals data for his own personal research. The book could have also explored topics more throughly, as some chapters felt rushed to the point they were briefly explained, then Pentland would conclude with a short description that his data proved otherwise based not a lot of actually data in the chapter but because he said so.
My complaint with the book is that near the end, it felt like a doctorate thesis about looser privacy regulations so he can get better access to individuals data for his own personal research. The book could have also explored topics more throughly, as some chapters felt rushed to the point they were briefly explained, then Pentland would conclude with a short description that his data proved otherwise based not a lot of actually data in the chapter but because he said so.
February 8, 2017
Sociale Big Data / Social Physics
Big Data is hot, gedrag is fascinerend. Combineer de twee en je hebt dit intrigerende boek van Alex Pentland: Sociale Big Data. Als professor en onderzoeker aan het MIT heeft hij zijn proefpersonen uitgerust met ‘sociometers’ die al hun gedrag (maar dan ook echt àlles) meten. Niet alleen kan hij met de resultaten gedrag voorspellen (predictive analytics), maar ook hoe gedrag zich verspreidt, als de griep zeg maar. Sterker nog: hij kan de griep voorspellen door ons veranderend gedrag! Ja ja, en je privacy dan? dacht ik toen ik dit las. Maar ook daar heeft hij een antwoord op in dit bijna wetenschappelijke boek dat veel details geeft, tot aan de algoritmes aan toe.
Waar komen nieuwe ideeën vandaan? En hoe worden die ideeën in actie omgezet, wordt het concrete innovatie? Dat is het werkgebied van de sociale fysica: de wiskundige verbanden tussen de ideeënstroom enerzijds en het gedrag van mensen anderzijds. Klinkt droog, maar dat is het bepaald niet. De data die worden gebruikt zijn telefoongegevens, pinbetalingen, gps-gegevens, enzovoorts. Dit is reality mining waarvoor Pentland ‘levende laboratoria’ gebruikt: groepen mensen worden van seconde tot seconde gevolgd, jaren achtereen. Een badge, de zogenoemde sociometer, verzamelt elke 16 miliseconde data over locatie, communicatie, lichaamstaal, spreektoon, wie er in de buurt is, workflow, taken van de drager en nog veel meer. Dat zou ik weleens willen zien!
Mensen beïnvloeden elkaar, inspireren elkaar. Met hoe meer mensen je omgaat buiten je directe omgeving (exploratie), hoe meer nieuwe ideeën je zult opdoen, en weer zult delen met anderen. Dit levert jou en je omgeving meer succes op. Deze relatie heeft Pentland aangetoond in diverse onderzoeken. Interessant is dat alléén maar omgaan met “vreemden” ook niet goed is, omdat je succes bereikt door samen te werken, en dat doe je met je naaste omgeving. Door alleen om te gaan met je naaste omgeving, kom je in een ‘echokamer’ terecht, waar iedereen alleen elkaars meningen herhaalt. Dodelijk voor innovatie. Er moet dus een balans komen tussen ‘exploratie’ en betrokkenheid. Pentland meet die balans door reality mining in organisaties, laat er wat algoritmes op los, en kan daardoor redelijk nauwkeurig succes en innovatie voorspellen.
Op grotere schaal worden deze voorspellende algoritmes gebruikt om bijvoorbeeld epidemieën te voorspellen. Als we ons niet zo lekker voelen verandert ons gedrag: we gaan meer met elkaar bellen. Als een hele woonwijk of stad dit doet, kan dit dus een hele vroege waarschuwing zijn voor een epidemie. Super nuttig lijkt me.
Als GRC-professional en IT-auditor heb ik natuurlijk direct vragen over privacy: hoe beschermen we al die data? En wie is daar eigenaar van? Pentland doet zijn initiatief ‘Data for Development’ of D4D uit de doeken, gestart in 2013 met een pilot over belpatronen en mobiliteit in Ivoorkust. Hiermee werd armoede en etnisch geweld onderzocht. In de datapool worden de data bewerkt met algoritmen en alleen de geaggregeerde resultaten worden verstrekt, niet de data zelf. Hierbij werd een juridisch contract gebruikt, waarin het doel en gebruik van de data precies is omschreven. Pentland voorziet een dergelijke structuur, OpenPDS, voor alle dataverzamelingen, wereldwijd. Het individu is eigenaar en bepaalt voor welk doel hij zijn data ter beschikking stelt.
De mogelijke nieuwe toepassingen van Big Data, en hoe dit de gezondheid in de wereld kan bevorderen en armoede kan bestrijden, vond ik erg inspirerend. Door de gedetailleerde uitwerking van de onderzoeken raakte ik er ook van overtuigd dat dit allemaal haalbaar is. Inhoudelijk is dit boek top!
Ik had wel wat moeite met de schrijfstijl. De auteur is een autoriteit op dit gebied en dat laat hij je weten ook. “Dit boek ……..is gebaseerd op een aantal van mijn artikelen die recentelijk zijn verschenen in ‘s werelds meest vooraanstaande wetenschappelijke tijdschriften”. Op elke bladzijde is een dergelijk “ik-statement” te vinden. In het begin verleent het wel de nodige geloofwaardigheid aan zijn stellingen, maar door de overdaad vond ik het al snel irritant worden. Amerikaanse stijl opschepperij die in Nederland minder goed valt, denk ik.
Daarnaast is het geschreven in een wat pompeuze stijl en zit er redelijk wat jargon in, als een wetenschappelijk artikel. Daardoor moet je heel geconcentreerd lezen. “Omdat de sociale fysica geen poging doet innerlijke cognitieve processen vast te leggen, is ze inherent probabilistisch, met een onreduceerbare kern van onzekerheid ten gevolge van het buiten beschouwing laten van de generatieve aard van het bewuste menselijke denken.” Zomaar een voorbeeld.
De grafieken zijn oorspronkelijk in kleur, maar nu in 50-tinten-grijs afgedrukt, waardoor de interpretatie best lastig is. Tenslotte is de vertaling soms wat onlogisch: de eerdergenoemde sociometer wordt vertaald als “naamkaartje” en ik realiseerde me pas later dat waarschijnlijk een badge wordt bedoeld. We kennen die al met chips voor toegang tot kantoren en de centrale printer, maar ook met tegoeden voor de betaling in de personeelskantine. Pentland’s badge is uitgerust met nog veel meer functionaliteit om beweging, audio en gedrag in het algemeen te registreren. Om zo’n badge met naamkaart te vertalen…..tja.
Als je door deze uiterlijkheden heen kunt prikken, heb je een bijzonder interessant boek in handen! 3 sterren ***
Sociale Big Data van Alex Pentland, uitgegeven bij Maven. ISBN 978 94 9184 533 8. Maven heeft mij een gratis boek toegestuurd voor deze recensie.
Big Data is hot, gedrag is fascinerend. Combineer de twee en je hebt dit intrigerende boek van Alex Pentland: Sociale Big Data. Als professor en onderzoeker aan het MIT heeft hij zijn proefpersonen uitgerust met ‘sociometers’ die al hun gedrag (maar dan ook echt àlles) meten. Niet alleen kan hij met de resultaten gedrag voorspellen (predictive analytics), maar ook hoe gedrag zich verspreidt, als de griep zeg maar. Sterker nog: hij kan de griep voorspellen door ons veranderend gedrag! Ja ja, en je privacy dan? dacht ik toen ik dit las. Maar ook daar heeft hij een antwoord op in dit bijna wetenschappelijke boek dat veel details geeft, tot aan de algoritmes aan toe.
Waar komen nieuwe ideeën vandaan? En hoe worden die ideeën in actie omgezet, wordt het concrete innovatie? Dat is het werkgebied van de sociale fysica: de wiskundige verbanden tussen de ideeënstroom enerzijds en het gedrag van mensen anderzijds. Klinkt droog, maar dat is het bepaald niet. De data die worden gebruikt zijn telefoongegevens, pinbetalingen, gps-gegevens, enzovoorts. Dit is reality mining waarvoor Pentland ‘levende laboratoria’ gebruikt: groepen mensen worden van seconde tot seconde gevolgd, jaren achtereen. Een badge, de zogenoemde sociometer, verzamelt elke 16 miliseconde data over locatie, communicatie, lichaamstaal, spreektoon, wie er in de buurt is, workflow, taken van de drager en nog veel meer. Dat zou ik weleens willen zien!
Mensen beïnvloeden elkaar, inspireren elkaar. Met hoe meer mensen je omgaat buiten je directe omgeving (exploratie), hoe meer nieuwe ideeën je zult opdoen, en weer zult delen met anderen. Dit levert jou en je omgeving meer succes op. Deze relatie heeft Pentland aangetoond in diverse onderzoeken. Interessant is dat alléén maar omgaan met “vreemden” ook niet goed is, omdat je succes bereikt door samen te werken, en dat doe je met je naaste omgeving. Door alleen om te gaan met je naaste omgeving, kom je in een ‘echokamer’ terecht, waar iedereen alleen elkaars meningen herhaalt. Dodelijk voor innovatie. Er moet dus een balans komen tussen ‘exploratie’ en betrokkenheid. Pentland meet die balans door reality mining in organisaties, laat er wat algoritmes op los, en kan daardoor redelijk nauwkeurig succes en innovatie voorspellen.
Op grotere schaal worden deze voorspellende algoritmes gebruikt om bijvoorbeeld epidemieën te voorspellen. Als we ons niet zo lekker voelen verandert ons gedrag: we gaan meer met elkaar bellen. Als een hele woonwijk of stad dit doet, kan dit dus een hele vroege waarschuwing zijn voor een epidemie. Super nuttig lijkt me.
Als GRC-professional en IT-auditor heb ik natuurlijk direct vragen over privacy: hoe beschermen we al die data? En wie is daar eigenaar van? Pentland doet zijn initiatief ‘Data for Development’ of D4D uit de doeken, gestart in 2013 met een pilot over belpatronen en mobiliteit in Ivoorkust. Hiermee werd armoede en etnisch geweld onderzocht. In de datapool worden de data bewerkt met algoritmen en alleen de geaggregeerde resultaten worden verstrekt, niet de data zelf. Hierbij werd een juridisch contract gebruikt, waarin het doel en gebruik van de data precies is omschreven. Pentland voorziet een dergelijke structuur, OpenPDS, voor alle dataverzamelingen, wereldwijd. Het individu is eigenaar en bepaalt voor welk doel hij zijn data ter beschikking stelt.
De mogelijke nieuwe toepassingen van Big Data, en hoe dit de gezondheid in de wereld kan bevorderen en armoede kan bestrijden, vond ik erg inspirerend. Door de gedetailleerde uitwerking van de onderzoeken raakte ik er ook van overtuigd dat dit allemaal haalbaar is. Inhoudelijk is dit boek top!
Ik had wel wat moeite met de schrijfstijl. De auteur is een autoriteit op dit gebied en dat laat hij je weten ook. “Dit boek ……..is gebaseerd op een aantal van mijn artikelen die recentelijk zijn verschenen in ‘s werelds meest vooraanstaande wetenschappelijke tijdschriften”. Op elke bladzijde is een dergelijk “ik-statement” te vinden. In het begin verleent het wel de nodige geloofwaardigheid aan zijn stellingen, maar door de overdaad vond ik het al snel irritant worden. Amerikaanse stijl opschepperij die in Nederland minder goed valt, denk ik.
Daarnaast is het geschreven in een wat pompeuze stijl en zit er redelijk wat jargon in, als een wetenschappelijk artikel. Daardoor moet je heel geconcentreerd lezen. “Omdat de sociale fysica geen poging doet innerlijke cognitieve processen vast te leggen, is ze inherent probabilistisch, met een onreduceerbare kern van onzekerheid ten gevolge van het buiten beschouwing laten van de generatieve aard van het bewuste menselijke denken.” Zomaar een voorbeeld.
De grafieken zijn oorspronkelijk in kleur, maar nu in 50-tinten-grijs afgedrukt, waardoor de interpretatie best lastig is. Tenslotte is de vertaling soms wat onlogisch: de eerdergenoemde sociometer wordt vertaald als “naamkaartje” en ik realiseerde me pas later dat waarschijnlijk een badge wordt bedoeld. We kennen die al met chips voor toegang tot kantoren en de centrale printer, maar ook met tegoeden voor de betaling in de personeelskantine. Pentland’s badge is uitgerust met nog veel meer functionaliteit om beweging, audio en gedrag in het algemeen te registreren. Om zo’n badge met naamkaart te vertalen…..tja.
Als je door deze uiterlijkheden heen kunt prikken, heb je een bijzonder interessant boek in handen! 3 sterren ***
Sociale Big Data van Alex Pentland, uitgegeven bij Maven. ISBN 978 94 9184 533 8. Maven heeft mij een gratis boek toegestuurd voor deze recensie.
February 20, 2015
One of the best books on social physics. It follows, using big data, the ways in witch not only good ideas are created but also how societies develop. Its main objective is to ask what are the roles and evolution of diversity, engagement, social trust, social intelligence and innovation, following both business management and urban development. A must read for those who want to be up to speed on the subject.
December 2, 2015
Alex Pentland is a numbers guy, and this book represents his initial attempts to parlaying to a lay readership his myriad published studies, summaries, and academic analyses of big data-derived computer models. The book is subdivided into three sections, but near as I can tell, there are really only two big ideas here: (1) that aggregated data can be reduced or distilled in such a way to permit "tuning" of human behavior without observer effects (akin to Asimov's Second Foundation and every economist's and dictator's dream) and (2) a cross between an explication of a data bill of rights and a paean to privacy that Pentland calls his "New Deal on Data."
Let's start with the extremely ambitious claim that forms the first big idea. What is "Social Physics," exactly, and how is it distinct from social psychology, economics, etc.? At pages 60-61, the author finds the concept rooted in folklore.
His first main example considers a Swiss energy conservation campaign (page 71).
Another of Pentland's experiments resonated better with me by combining feedback with incentives in a novel way. His team found that dieters were more likely to initiate and sustain their weight loss programs when incentivized by a peer group who encouraged them than by any direct reward (see, e.g., the success of social programs like Weight Watchers and "The Biggest Loser"). Not that financial or material incentives are misplaced, just that the trick was to offer them up to a subject's peers to promote their participation rather than directly to the subject. (It takes a village.) Yet I couldn't make head or tail of the method the author described using to quantify these conclusions.
Perhaps it appears in the wide range of papers Pentland and his co-authors have published? Because the author provides URL links to these, as well as to his data sets, and computer implementations of the various equations, the reader needn't just take his word for it, but then, Social Physics doesn't do that great a job in communicating his ideas of findings to a lay audience. The passage quoted strikes me as a pretty dense way of articulating parametric guesswork, but I guess the ultimate proof of the model lies in the pudding from which it was derived. How consistently does it match patterns observed in the data set? Do the elements of other data sets evincing similar patterns correlate with one another well, or is this more a case of fitting various data to extract a preexisting curve? Pentland does not seem to address these questions, nor to dispel possible counterexplanations for his various conclusions.
"The sociometric data that my research group and I have gathered from many different organizations," he claims at page 96, "show that creative output depends strongly on two processes: idea discovery (exploration) and the integration of those ideas into new behaviors (engagement)." The rate of idea flow is the metric by which he determines a group's ability to circulate new information and modify behavior accordingly and that group effectiveness can be maximized by tuning it to ideal rates. If this is so (again assuming I understand him correctly), his metric would seem a powerful tool for improving performance and social cohesion. However, I never was able to understand exactly how this should be operationalized: what does one do to "tune" the idea flow of a given social system? Does it vary on a case-by-case basis? What other explanations are there for group performance and behavior? His examples don't seem to account for the effects of independent variables, but perhaps it doesn't matter. Just fiddle with seating arrangements, schedules, work assignments, and performance incentives until you have it right. Who needs to triage precise causes and effects?
For example, Pentland's studies of select help desks and call centers revealed that individual productivity increased materially over the mean output from people kept seatbelted to their desk phones, noses to the grind-monitor when employees were allowed and encouraged to socialize with one another at regular breaks. The author attributed this improved performance to increased idea flow among the staff ("being in the loop allows employees to learn tricks of the trade"). Then again, people are social animals, so perhaps some of the credit should go to improved morale. Pentland does not consider or distinguish the alternative.
The models he describes for cities bear marked similarity to the sort of ideal urban planning described in the 1970's by Christopher Alexander in A Pattern Language, and why shouldn't they? Each are derived not just from theories about human psychology, but extensive trial-and-error observations of human interaction. To say nothing of Bill Gore's Rule of 150, which Malcolm Gladwell recounted in The Tipping Point. Thus, at page 168, "We want to increase engagement in the residential areas, which will lead to stronger norms of behavior, but not increase the amount of exploration for everyone, since that would lead to more crime along with greater innovation…. The best size for such a city can even be calculated: if within each peer group everyone is a friend of a friend, then the math of social physics indicates that we get maximum engagement for populations of up to roughly one hundred thousand people. This suggests that the best solution is small-to-medium-sized towns in which everyone is within walking distance of a town center, the stores, the schools, and the clinics."
The author closes out the book promoting his "New Deal on Data," and it's not hard to see why.
Although I applaud his efforts to establish a legal regime that might at least afford post hoc remedies to harmed individuals, I'm personally skeptical that its adoption does anything to protect privacy or keep us from becoming some future emperor's puppet-like Borglings. Welcome to the new age. See Asimov, Isaac, again, and a happy goldfish bowl to all.
Let's start with the extremely ambitious claim that forms the first big idea. What is "Social Physics," exactly, and how is it distinct from social psychology, economics, etc.? At pages 60-61, the author finds the concept rooted in folklore.
The idea of a collective intelligence that develops within communities is an old one; indeed, it is embedded in the English language. Consider the word "kith," familiar to modern English speakers from the phrase "kith and kin." Derived from old English and old German words for knowledge, kith refers to a more or less cohesive group with common beliefs and customs. These are also the roots for "couth," which means possessing a high degree of sophistication, as well as its more familiar counterpart, "uncouth." Thus, our kith is the circle of peers (not just friends) from whom we learn the "correct" habits of action…. We learn common sense almost automatically, by observing and then copying the common behaviors of our peers.The whole is presented with a great deal of confidence, and I am thus cowed into thinking he may actually know what he's talking about. That said, I found this book extremely slow going and am not sure I come away from it any better enlightened. Manipulations of statistical models aside, Pentland is not the best communicator.
His first main example considers a Swiss energy conservation campaign (page 71).
In the first experiment, home owners received social feedback on how much electricity they used relative to the average person. When the comparison was between the home owner and all other people in their country, virtually no savings resulted; people behaved the same. When the comparison was between them and people in their neighborhood, however, things worked better, showing that how closely they identified with the people in the comparison group mattered. This is a social network effect: Identification with a group of people increases both trust of group members and the social pressure that the group can exert.All praise social physics? I wasn't fully persuaded by this example. Another explanation for the difference could be that people receiving the local comparison found the feedback more immediately relevant to them than data from a more distant and less personal environment. In other words, rather than indicating a social network effect per se, the campaign just showed that like a polished mirror, feedback works better the more accurately it appears to reflect its subject.
Another of Pentland's experiments resonated better with me by combining feedback with incentives in a novel way. His team found that dieters were more likely to initiate and sustain their weight loss programs when incentivized by a peer group who encouraged them than by any direct reward (see, e.g., the success of social programs like Weight Watchers and "The Biggest Loser"). Not that financial or material incentives are misplaced, just that the trick was to offer them up to a subject's peers to promote their participation rather than directly to the subject. (It takes a village.) Yet I couldn't make head or tail of the method the author described using to quantify these conclusions.
The influence model breaks this overall "company state" into the influence each person c has on a particular other person c'... where the influence matrix, R^c', ^c, captures the influence strength of person c over c' and describes how influence spreads through the company's social network. The number of parameters in this model grows relatively slowly with increasing numbers of people and their internal states, making it easy to mathematically model "live" data and use it in real-time applications. Practically, this means we can determine the influence model parameters -- influence, states, etc. -- without knowing the social ties or learned behaviors beforehand by using an expectation maximization algorithm…. For almost all of the examples in this book, including the role of social influence on political views, purchasing behavior, and health choices, as well as productivity in small groups, departments within companies, and entire cities, we find that using measures of the amount of social interaction -- both direct and indirect -- in order to estimate social influence produces accurate estimates of future behavior. (pages 82-3)Say what now? Leaving aside the details of Pentland's model, if I read that last sentence correctly, it would seem that he expects to move from general principles of the sort that lend greater post hoc understanding to specific behavioral predictions of the sort that might be deliberately manipulated. That's surely the holy grail of public policy, but if that's what he intends to pull off, I'm at a loss to find any place where this assertion is clearly demonstrated or exemplified in his book.
Perhaps it appears in the wide range of papers Pentland and his co-authors have published? Because the author provides URL links to these, as well as to his data sets, and computer implementations of the various equations, the reader needn't just take his word for it, but then, Social Physics doesn't do that great a job in communicating his ideas of findings to a lay audience. The passage quoted strikes me as a pretty dense way of articulating parametric guesswork, but I guess the ultimate proof of the model lies in the pudding from which it was derived. How consistently does it match patterns observed in the data set? Do the elements of other data sets evincing similar patterns correlate with one another well, or is this more a case of fitting various data to extract a preexisting curve? Pentland does not seem to address these questions, nor to dispel possible counterexplanations for his various conclusions.
"The sociometric data that my research group and I have gathered from many different organizations," he claims at page 96, "show that creative output depends strongly on two processes: idea discovery (exploration) and the integration of those ideas into new behaviors (engagement)." The rate of idea flow is the metric by which he determines a group's ability to circulate new information and modify behavior accordingly and that group effectiveness can be maximized by tuning it to ideal rates. If this is so (again assuming I understand him correctly), his metric would seem a powerful tool for improving performance and social cohesion. However, I never was able to understand exactly how this should be operationalized: what does one do to "tune" the idea flow of a given social system? Does it vary on a case-by-case basis? What other explanations are there for group performance and behavior? His examples don't seem to account for the effects of independent variables, but perhaps it doesn't matter. Just fiddle with seating arrangements, schedules, work assignments, and performance incentives until you have it right. Who needs to triage precise causes and effects?
For example, Pentland's studies of select help desks and call centers revealed that individual productivity increased materially over the mean output from people kept seatbelted to their desk phones, noses to the grind-monitor when employees were allowed and encouraged to socialize with one another at regular breaks. The author attributed this improved performance to increased idea flow among the staff ("being in the loop allows employees to learn tricks of the trade"). Then again, people are social animals, so perhaps some of the credit should go to improved morale. Pentland does not consider or distinguish the alternative.
The models he describes for cities bear marked similarity to the sort of ideal urban planning described in the 1970's by Christopher Alexander in A Pattern Language, and why shouldn't they? Each are derived not just from theories about human psychology, but extensive trial-and-error observations of human interaction. To say nothing of Bill Gore's Rule of 150, which Malcolm Gladwell recounted in The Tipping Point. Thus, at page 168, "We want to increase engagement in the residential areas, which will lead to stronger norms of behavior, but not increase the amount of exploration for everyone, since that would lead to more crime along with greater innovation…. The best size for such a city can even be calculated: if within each peer group everyone is a friend of a friend, then the math of social physics indicates that we get maximum engagement for populations of up to roughly one hundred thousand people. This suggests that the best solution is small-to-medium-sized towns in which everyone is within walking distance of a town center, the stores, the schools, and the clinics."
The author closes out the book promoting his "New Deal on Data," and it's not hard to see why.
[T]he most important generator of city data is a familiar tool: the ubiquitous mobile phone. These devices are, in effect, personal sensing devices that are becoming more powerful and more sophisticated with each product iteration…. [W]e can… gauge the mood of a crowd by analyzing the digital chatter…. Consumers are also beginning to make purchases simply by scanning items with their phones, thereby adding financial and product choice information to [real-time captures of biometric data]. (page 138)What's in *your* wallet? Clearly the ability to undetectably shape human behavior at any scale is a power that can be used for good or ill, the difference being a mere matter of perspective. Pentland would like such godlike omniscience to be used beneficently, so having opened what he calls the "Promethean Fire" of digital transparency, it should be no surprise that he feels a concomitant burden to promote some norms. He advocates for an international treaty that would require all the various data streams that form the smartphone's inherent human-to-internet interface be anonymized in the aggregate, made universally accessible in open source repositories, while somehow being personally accessible, readable, and controllable by individuals (in that all uses must be opt-in and any personal information could be personally destroyed).
Although I applaud his efforts to establish a legal regime that might at least afford post hoc remedies to harmed individuals, I'm personally skeptical that its adoption does anything to protect privacy or keep us from becoming some future emperor's puppet-like Borglings. Welcome to the new age. See Asimov, Isaac, again, and a happy goldfish bowl to all.
October 15, 2014
Big data is inmiddels een begrip geworden dat vrijwel iedereen bekend in de oren klinkt. Zeker nadat voormalig medewerker van de CIA, Edward Snowden, de hele wereld vertelde over PRISM, het programma dat door de National Security Agency (NSA) wordt gebruikt om inlichtingen te verkrijgen uit de digitale gegevens die gebruikers van software en apparatuur van Microsoft, Yahoo!, Google, Facebook, YouTube, Skype, Apple, e.a. achterlaten.
Ongeveer tegelijkertijd kwam het boek De Big Data-revolutie van Viktor Mayer-Schönberger, Kenneth Cukier uit, waarin niet alleen aandacht werd besteed aan de nadelen van het gebruik van Big Data, maar ook aan de voordelen. Wat Viktor Mayer-Schönberger en Kenneth Cukier vooral benadrukken is dat er haast moet worden gemaakt met het aanpassen van de privacywetgeving en de rechtspraak, omdat de huidig regels en wetten niet meer voldoen.
Alex Pentland, de schrijver van Sociale Big Data, is het daarmee volkomen mee eens. In het hoofdstuk over Datagedreven samenleveningen, schrijft hij bijvoorbeeld
Een belangrijke reden om dit nieuwe beleid op te stellen is dat onze data meer waard zijn wanneer ze worden gedeeld, omdat ze de basis kunnen vormen voor verbeteringen in het systeem van de volksgezondheid, het vervoer en het openbaar bestuur [...]
Een goed voorbeeld daarvan is Google Flu, dat zowel door Mayer-Schönberger en Cukier als Pentland wordt genoemd. Google legt uit dat ze "hebben ontdekt dat er een sterk verband is tussen het aantal mensen dat zoekt naar grieponderwerpen en het aantal mensen dat daadwerkelijk griepsymptomen heeft. Natuurlijk is niet iedereen die zoekt naar 'griep' zelf ziek, maar als alle griepgerelateerde zoekopdrachten samen worden bekeken, wordt er toch een patroon zichtbaar.[...] Door te analyseren hoe vaak deze zoekopdrachten worden uitgevoerd, kunnen we een schatting maken van de verspreiding van griep in verschillende landen en regio's wereldwijd."
Pentland stelt vervolgens dat als de verspreiding kan worden voorspeld, deze vanzelfsprekend ook een halt kan worden toegeroepen, waarmee het gevaar van een pandemie veel kleiner is geworden. Helaas voor Pentland (en Mayer-Schönberger en Cukier) werd begin dit jaar bekend dat Google Flu Trend de verspreiding van de griep lang niet zo goed meer voorspelt als in 2009 en 2010. De oorzaak daarvan is moeilijk te vinden aangezien Google niet verteld welke zoektermen ze gebruiken en op welke manier ze tot hun conclusies komen. Waarmee, zijns ondanks, wordt aangetoond dat het pleidooi van Pentland voor verandering van de privacywetgeving en data-eigendom van groot belang is, omdat controle van gebruik en analyse-uitkomsten van cruciaal belang is.
Wat Pentland echter met dit boek vooral wil bereiken is dat de top van de overheid en het bedrijfsleven afstapt van het uitsluitende individugerichte economische en politieke denken en daar sociale interacties bij betrekt. Hij laat zien dat sociaal leren en sociale druk de primaire krachten zijn die veranderingen in cultuur bewerkstelligen en dat je door daarop in te grijpen de veranderingen kunt (bij)sturen.Zo liet de Bank of America de medewerkers van hun enorme call center één voor één koffiedrinken omdat dat 'gepraat met elkaar' toch nergens toe diende; de scripts vertelden tenslotte precies wat er gevraagd en gedaan moest worden. Het was daarom efficiënter om de medewerkers niet alleen op hun werkplek gescheiden te houden, maar ook tijdens de diverse pauzes.Toen echter, op basis van aanbevelingen van Pentlands team besloten werd om alle leden van een team gelijktijdig pauze te gunnen, schoot de productiviteit van het call center met 15 % omhoog. Mensen leren van elkaar, ook in pauzes. Of misschien wel juist in pauzes. Bij het koffiezetapparaat.
Pentland bespreekt in het boek diverse voorbeelden, legt daarbij uit op welke wijze de gegevens werden vergaard, wat die gegevens vertelden en wat er op basis van die gegevens geconcludeerd werd. En welke gevolgen een aanpassing had. Soms, overigens, was de aanpassing niet meer dan laten zien welke sociale interacties tot dan toe hadden plaatsgevonden. Zien dat de afdeling klantenservice totaal geïsoleerd is en dus niet op de hoogte van de veranderingen en nieuwigheden, zorgt ervoor dat interactie wordt aangepast door de marketingafdeling. Dat is al prikkel genoeg.
In de appendix Wiskunde wordt het het model besproken waarmee de invloed, het sociale leren en de sociale druk tussen individuen in een systeem gemodelleerd wordt. Het is het enige deel van het boek waarin veel aandacht is voor de wiskundige onderbouwing van Pentlands theorieën. Hij streeft er in de rest van het boek naar om de wiskundige formules handen en voeten te geven door duidelijke uitleg van de door hem gebruikte termen als ideeënstroom, exploratie, betrokkenheid, echokamers, sociale druk, sociaal leren en collectieve intelligentie, maar vooral door een fiks aantal prima voorbeelden die laten zien wat er met deze termen bedoeld wordt. Hij laat bijvoorbeeld zien hoe je ervoor kunt zorgen dat de creativiteit en productiviteit van een stad omhoog gaat zonder dat de criminaliteit ook omhoog gaat. Met andere woorden, hoe ervoor gezorgd kan worden dat er een dorpse sfeer is (betrokkenheid), maar de inwoners wel veel en vaak met nieuwe ideeën in aanraking komen (exploratie) waardoor de creativiteit en productiviteit stijgen.
Zoals gezegd wil Pentland wereldleiders, mensen in de top van overheden, multinationals en andere bedrijven vertellen dat er meer is dan markten, individuen en concurrentie. Eigenlijk wil hij ze - opnieuw - laten laten luisteren naar Adam Smith en hoopt hij dat dit boek daaraan bijdraagt. Ik hoop dat met hem. Aan de kwaliteit van vertellen en de begrijpelijkheid zal het niet liggen.
It is human nature to exchange not only goods, but also ideas, assistance and favors out of sympathy. It is these exchanges that guide men to create solutions for the good of the community.
Adam Smith.
Zie ook:
- VPRO Tegenlicht: Alex Pentland
- PRISM (wikipedia)
- Google.org Flutrends
- Google Flu Trends is no longer good at predicting flu, scientists find of Google Flu Trends (wikipedia)
- Dossier Big Data van The Guardian (verzameling Engelstalige big data verhalen)
- Reinventing society in the wake of Big Data
- Sandy Pentland: "Social Physics: How Good Ideas Spread" (Talks at Google)
Recensie geschreven voor Hebban | Wereld & Opinie
Ongeveer tegelijkertijd kwam het boek De Big Data-revolutie van Viktor Mayer-Schönberger, Kenneth Cukier uit, waarin niet alleen aandacht werd besteed aan de nadelen van het gebruik van Big Data, maar ook aan de voordelen. Wat Viktor Mayer-Schönberger en Kenneth Cukier vooral benadrukken is dat er haast moet worden gemaakt met het aanpassen van de privacywetgeving en de rechtspraak, omdat de huidig regels en wetten niet meer voldoen.
Alex Pentland, de schrijver van Sociale Big Data, is het daarmee volkomen mee eens. In het hoofdstuk over Datagedreven samenleveningen, schrijft hij bijvoorbeeld
Een belangrijke reden om dit nieuwe beleid op te stellen is dat onze data meer waard zijn wanneer ze worden gedeeld, omdat ze de basis kunnen vormen voor verbeteringen in het systeem van de volksgezondheid, het vervoer en het openbaar bestuur [...]
Een goed voorbeeld daarvan is Google Flu, dat zowel door Mayer-Schönberger en Cukier als Pentland wordt genoemd. Google legt uit dat ze "hebben ontdekt dat er een sterk verband is tussen het aantal mensen dat zoekt naar grieponderwerpen en het aantal mensen dat daadwerkelijk griepsymptomen heeft. Natuurlijk is niet iedereen die zoekt naar 'griep' zelf ziek, maar als alle griepgerelateerde zoekopdrachten samen worden bekeken, wordt er toch een patroon zichtbaar.[...] Door te analyseren hoe vaak deze zoekopdrachten worden uitgevoerd, kunnen we een schatting maken van de verspreiding van griep in verschillende landen en regio's wereldwijd."
Pentland stelt vervolgens dat als de verspreiding kan worden voorspeld, deze vanzelfsprekend ook een halt kan worden toegeroepen, waarmee het gevaar van een pandemie veel kleiner is geworden. Helaas voor Pentland (en Mayer-Schönberger en Cukier) werd begin dit jaar bekend dat Google Flu Trend de verspreiding van de griep lang niet zo goed meer voorspelt als in 2009 en 2010. De oorzaak daarvan is moeilijk te vinden aangezien Google niet verteld welke zoektermen ze gebruiken en op welke manier ze tot hun conclusies komen. Waarmee, zijns ondanks, wordt aangetoond dat het pleidooi van Pentland voor verandering van de privacywetgeving en data-eigendom van groot belang is, omdat controle van gebruik en analyse-uitkomsten van cruciaal belang is.
Wat Pentland echter met dit boek vooral wil bereiken is dat de top van de overheid en het bedrijfsleven afstapt van het uitsluitende individugerichte economische en politieke denken en daar sociale interacties bij betrekt. Hij laat zien dat sociaal leren en sociale druk de primaire krachten zijn die veranderingen in cultuur bewerkstelligen en dat je door daarop in te grijpen de veranderingen kunt (bij)sturen.Zo liet de Bank of America de medewerkers van hun enorme call center één voor één koffiedrinken omdat dat 'gepraat met elkaar' toch nergens toe diende; de scripts vertelden tenslotte precies wat er gevraagd en gedaan moest worden. Het was daarom efficiënter om de medewerkers niet alleen op hun werkplek gescheiden te houden, maar ook tijdens de diverse pauzes.Toen echter, op basis van aanbevelingen van Pentlands team besloten werd om alle leden van een team gelijktijdig pauze te gunnen, schoot de productiviteit van het call center met 15 % omhoog. Mensen leren van elkaar, ook in pauzes. Of misschien wel juist in pauzes. Bij het koffiezetapparaat.
Pentland bespreekt in het boek diverse voorbeelden, legt daarbij uit op welke wijze de gegevens werden vergaard, wat die gegevens vertelden en wat er op basis van die gegevens geconcludeerd werd. En welke gevolgen een aanpassing had. Soms, overigens, was de aanpassing niet meer dan laten zien welke sociale interacties tot dan toe hadden plaatsgevonden. Zien dat de afdeling klantenservice totaal geïsoleerd is en dus niet op de hoogte van de veranderingen en nieuwigheden, zorgt ervoor dat interactie wordt aangepast door de marketingafdeling. Dat is al prikkel genoeg.
In de appendix Wiskunde wordt het het model besproken waarmee de invloed, het sociale leren en de sociale druk tussen individuen in een systeem gemodelleerd wordt. Het is het enige deel van het boek waarin veel aandacht is voor de wiskundige onderbouwing van Pentlands theorieën. Hij streeft er in de rest van het boek naar om de wiskundige formules handen en voeten te geven door duidelijke uitleg van de door hem gebruikte termen als ideeënstroom, exploratie, betrokkenheid, echokamers, sociale druk, sociaal leren en collectieve intelligentie, maar vooral door een fiks aantal prima voorbeelden die laten zien wat er met deze termen bedoeld wordt. Hij laat bijvoorbeeld zien hoe je ervoor kunt zorgen dat de creativiteit en productiviteit van een stad omhoog gaat zonder dat de criminaliteit ook omhoog gaat. Met andere woorden, hoe ervoor gezorgd kan worden dat er een dorpse sfeer is (betrokkenheid), maar de inwoners wel veel en vaak met nieuwe ideeën in aanraking komen (exploratie) waardoor de creativiteit en productiviteit stijgen.
Zoals gezegd wil Pentland wereldleiders, mensen in de top van overheden, multinationals en andere bedrijven vertellen dat er meer is dan markten, individuen en concurrentie. Eigenlijk wil hij ze - opnieuw - laten laten luisteren naar Adam Smith en hoopt hij dat dit boek daaraan bijdraagt. Ik hoop dat met hem. Aan de kwaliteit van vertellen en de begrijpelijkheid zal het niet liggen.
It is human nature to exchange not only goods, but also ideas, assistance and favors out of sympathy. It is these exchanges that guide men to create solutions for the good of the community.
Adam Smith.
Zie ook:
- VPRO Tegenlicht: Alex Pentland
- PRISM (wikipedia)
- Google.org Flutrends
- Google Flu Trends is no longer good at predicting flu, scientists find of Google Flu Trends (wikipedia)
- Dossier Big Data van The Guardian (verzameling Engelstalige big data verhalen)
- Reinventing society in the wake of Big Data
- Sandy Pentland: "Social Physics: How Good Ideas Spread" (Talks at Google)
Recensie geschreven voor Hebban | Wereld & Opinie
December 28, 2017
The case studies, by using big data, introduced in this book indicate that (a) people primarily rely on their peers for social learning and interactions (flow of ideas) hence shape highly patterned behavior based on “common sense”, (b) repeated cooperative interactions among members of the community (engagement) brings movement toward cooperative behavior, (c) to enhance face-to-face engagement is a key factor to improve productivity and (subjectively judged) creativity, and (d) social physics is applicable in urban planning so as to design city in such a way higher social tie density produces greater levels of idea flow, leading to increases in productivity and innovation.
From my viewpoint, social physics proposed by Pentland seems merely a sociological version of institutional economics. The concepts highlighted in this social physics such as engagement, trust, social pressure, social tie, and enforcement, had commonly used in transaction cost (or information cost) theory in institutional economics. The findings introduced from previous studies used social physics approach are also the same or similar ones already identified in previous studies of institutional economics more than two decades ago.
What something new must be data collection method, big data, which become usable for the analysis of a certain huge size of population, thank to the ubiquitous online networks in our daily life. For instance, the author and his PhD students used smartphones installed a special software and gave them the research target people to record who called, e-mailed, or texted whom, who were active friends on social media, who spent time face-to-face, and where they spent time, in their research of a community of young families. Such a way of data collection, almost real-time, the hundreds of gigabytes of over 1.5 million hours of automatically recorded data, would be technically impossible two decades ago before smartphone use became popular among ordinary people.
The other insight of the author expressed in this book is that once a social norm is in place through face-to-face interactions, then people can reinforce a trusted relationship, even though they remain physically separated. It seems important function of online social network to reshape communities among friends, colleagues and neighbors.
From my viewpoint, social physics proposed by Pentland seems merely a sociological version of institutional economics. The concepts highlighted in this social physics such as engagement, trust, social pressure, social tie, and enforcement, had commonly used in transaction cost (or information cost) theory in institutional economics. The findings introduced from previous studies used social physics approach are also the same or similar ones already identified in previous studies of institutional economics more than two decades ago.
What something new must be data collection method, big data, which become usable for the analysis of a certain huge size of population, thank to the ubiquitous online networks in our daily life. For instance, the author and his PhD students used smartphones installed a special software and gave them the research target people to record who called, e-mailed, or texted whom, who were active friends on social media, who spent time face-to-face, and where they spent time, in their research of a community of young families. Such a way of data collection, almost real-time, the hundreds of gigabytes of over 1.5 million hours of automatically recorded data, would be technically impossible two decades ago before smartphone use became popular among ordinary people.
The other insight of the author expressed in this book is that once a social norm is in place through face-to-face interactions, then people can reinforce a trusted relationship, even though they remain physically separated. It seems important function of online social network to reshape communities among friends, colleagues and neighbors.
September 1, 2023
Libro che contiene idee interessanti. Ma andiamo con ordine.
Perché fisica sociale? Gli studi dell’autore ci dicono che come la fisica con le sue leggi è capace di predire lo stato ed i movimenti della materia, così l’utilizzo della matematica e dei big data è capace di regolare i comportamenti sociali.
Siete spaventati da questi studi e sperimentazioni?
Lo stesso autore fa una premessa, per passare dalle sperimentazioni locali ad effetti globali è necessario che gli organi competenti producano leggi capaci di regolamentare la privacy dei cittadini. Quindi in un possibile assetto futuro il cittadina dovrà prendersi cura non solo della sua vita materiale, ma anche della gestione ed utilizzo dei propri dati digitali.
Detto questo il testo propone le seguenti idee di fondo.
Se nell’Ottocento si promuoveva l’individualismo e la realizzazione del se, in futuro potrebbe essere probabile una società basata sullo scambio economico e di idee, dove la cooperazione supera gli individualismi. Praticamente un cambio di paradigma.
Centrale in questa futura società è il flusso di idee. L’autore considera la circolazione delle idee basta su due principi di apprendimento quello sociale e quello individuale. L’apprendimento sociale si fa per emulazione quello individuale per studio e riflessione, questi due metodi di apprendimento possono essere collegati a due metodi di pensiero quello lento e quello veloce. Il pensiero veloce che è predominante in termini quantitativi si rifà all’apprendimento sociale, quello lento più raro riguarda l’apprendimento individuale.
Data la predominanza dell’apprendimento sociale, l’autore ci dice che avere società con un maggiore flusso di idee equivale ad avere società culturalmente avanzate.
Su questi basi teoriche è possibile utilizzare i big data per monitorare e costruire reti globali e locali con cui governare aziende città e nazioni.
Purtroppo da un grande potere derivano grandi responsabilità. Non è molto chiaro come il monitoraggio dei big data possa influenzare il nostro libero arbitrio; non è molto chiaro chi sarà a detenere le competenze per indirizzare la nostra società.
Una cosa è certe con l’utilizzo dei big data ci sono dei margini di miglioramento per la nostra società; ne vale la pena? È veramente la direzione giusta da prendere per l’umanità?
Perché fisica sociale? Gli studi dell’autore ci dicono che come la fisica con le sue leggi è capace di predire lo stato ed i movimenti della materia, così l’utilizzo della matematica e dei big data è capace di regolare i comportamenti sociali.
Siete spaventati da questi studi e sperimentazioni?
Lo stesso autore fa una premessa, per passare dalle sperimentazioni locali ad effetti globali è necessario che gli organi competenti producano leggi capaci di regolamentare la privacy dei cittadini. Quindi in un possibile assetto futuro il cittadina dovrà prendersi cura non solo della sua vita materiale, ma anche della gestione ed utilizzo dei propri dati digitali.
Detto questo il testo propone le seguenti idee di fondo.
Se nell’Ottocento si promuoveva l’individualismo e la realizzazione del se, in futuro potrebbe essere probabile una società basata sullo scambio economico e di idee, dove la cooperazione supera gli individualismi. Praticamente un cambio di paradigma.
Centrale in questa futura società è il flusso di idee. L’autore considera la circolazione delle idee basta su due principi di apprendimento quello sociale e quello individuale. L’apprendimento sociale si fa per emulazione quello individuale per studio e riflessione, questi due metodi di apprendimento possono essere collegati a due metodi di pensiero quello lento e quello veloce. Il pensiero veloce che è predominante in termini quantitativi si rifà all’apprendimento sociale, quello lento più raro riguarda l’apprendimento individuale.
Data la predominanza dell’apprendimento sociale, l’autore ci dice che avere società con un maggiore flusso di idee equivale ad avere società culturalmente avanzate.
Su questi basi teoriche è possibile utilizzare i big data per monitorare e costruire reti globali e locali con cui governare aziende città e nazioni.
Purtroppo da un grande potere derivano grandi responsabilità. Non è molto chiaro come il monitoraggio dei big data possa influenzare il nostro libero arbitrio; non è molto chiaro chi sarà a detenere le competenze per indirizzare la nostra società.
Una cosa è certe con l’utilizzo dei big data ci sono dei margini di miglioramento per la nostra società; ne vale la pena? È veramente la direzione giusta da prendere per l’umanità?
December 23, 2017
This is a very good read that I would consider a part of a growing collection of very similar "21st century interdisciplinary social science" themed, accessible reads. This is a read somewhat echoing works like Linked by Barabasi, Going Viral by Nahon and Hemsley, Triumph of the City by Glaeser, The Stack by Bratton, and others. Social Physics echoes a new brand of a 'new science' steeped in an era of big data science, modern statistics, and network methodologies with a focus on tying older paradigms of 'theory' oriented social science.
This book echoes similar perspectives that are stated in many networked minded books; it contains big ideas and questions for the future of scientific inquiry and society. It develops a moderately new way of thinking. It challenges the purist notions of our rationalist past to consider beyond our own, reportedly independent minds in order to develop a better understanding of the real structure of the world.
That being said, this is NOT a "new science". This is not "Physics". It markets a idea of a new "positivist" perspective, but it never actualizes that idea. Instead it promotes the epistemology of "statistical truths". This is somewhat misleading, and his language sometimes minces the argument and weasels a little of what is interpreted. This is not to say that the work he cites (the real interests of this "new science") is not worthy of consideration. In fact, very much the opposite. The vast interest here is in the significance of his current work. Pentland takes a lot of time to discuss the program he's developed from his research in an empirical sense. THIS is the new science; it is not the philosophy and body of theory it reports. It is through those studies that you see the significance.
All that is to say, if you are interested in in empirical developments, this is the book! If you want the philosophical aspects, read Barabasi. If you want urban city development research, read Glaeser. If you want social media methods and communication, read Hemsley and Nahon.
This book echoes similar perspectives that are stated in many networked minded books; it contains big ideas and questions for the future of scientific inquiry and society. It develops a moderately new way of thinking. It challenges the purist notions of our rationalist past to consider beyond our own, reportedly independent minds in order to develop a better understanding of the real structure of the world.
That being said, this is NOT a "new science". This is not "Physics". It markets a idea of a new "positivist" perspective, but it never actualizes that idea. Instead it promotes the epistemology of "statistical truths". This is somewhat misleading, and his language sometimes minces the argument and weasels a little of what is interpreted. This is not to say that the work he cites (the real interests of this "new science") is not worthy of consideration. In fact, very much the opposite. The vast interest here is in the significance of his current work. Pentland takes a lot of time to discuss the program he's developed from his research in an empirical sense. THIS is the new science; it is not the philosophy and body of theory it reports. It is through those studies that you see the significance.
All that is to say, if you are interested in in empirical developments, this is the book! If you want the philosophical aspects, read Barabasi. If you want urban city development research, read Glaeser. If you want social media methods and communication, read Hemsley and Nahon.
March 4, 2024
Interesting book and obviously written for a general audience. It was less definitive and substantive than I had hoped. Also, and I'm not sure whether the author noticed this, but the portrayal is clearly more aligned with environmental science or biology than physics. For example, when talking about misalignment in incentives and space configuration to allow social forces to create more experiences and connections, there was no discussion about the fact that every configuration change brings about new predatory mechanisms (think viruses, cancers, bacteria, etc).
Anyway, brief points:
* Social sciences have lacked a unifying fundamental framework for mathematically predicting the outcomes of interactions.
* The author's work has helped push these fields towards being more quantitative and predictable.
* Information exchange, points of contact, frequency of contact, network size, etc, were all predictive variables
* Social incentives, such as wanting to give a friend a 'free lottery ticket' (red balloon contest) drove engagement more than strict monetary incentives
* (In 'things that are not enws',) social media interactions were not predictive and did not fulfill the same purposes as in-person collaboration
* Information transfer was found to be highly influential in group decision-making, more so than other factors
* Hierarchies and power differentials squash information transfer (which we already knew from high power distance cultures causing airplane crashes)
* Infrastructure can have unintended social consequences (creativity, crime) but these consequences can be mitigated and/or controlled with a predictive framework in place (this aligns well with A Pattern Language on the subject of how to construct residential and business areas btw)
* Mass repositories of shared resources help everyone (for example, d4d.net)
* Limited information transfer creates aggressively antisocial cultures (the flipside of this point is covered in Crucial Conversations as contributions to the shared pool of meaning)
Anyway, brief points:
* Social sciences have lacked a unifying fundamental framework for mathematically predicting the outcomes of interactions.
* The author's work has helped push these fields towards being more quantitative and predictable.
* Information exchange, points of contact, frequency of contact, network size, etc, were all predictive variables
* Social incentives, such as wanting to give a friend a 'free lottery ticket' (red balloon contest) drove engagement more than strict monetary incentives
* (In 'things that are not enws',) social media interactions were not predictive and did not fulfill the same purposes as in-person collaboration
* Information transfer was found to be highly influential in group decision-making, more so than other factors
* Hierarchies and power differentials squash information transfer (which we already knew from high power distance cultures causing airplane crashes)
* Infrastructure can have unintended social consequences (creativity, crime) but these consequences can be mitigated and/or controlled with a predictive framework in place (this aligns well with A Pattern Language on the subject of how to construct residential and business areas btw)
* Mass repositories of shared resources help everyone (for example, d4d.net)
* Limited information transfer creates aggressively antisocial cultures (the flipside of this point is covered in Crucial Conversations as contributions to the shared pool of meaning)
Displaying 1 - 30 of 184 reviews