Complexity. The new world between chance and choice
April 10, 2010
Nonlinear dynamics are concerned with complex, messy systems. Examples for these systems are the human brain, the evolution of life itself and the weather. There is not a single science of non-linearity, but there are different streams of research such as chaos theory or the theory of complex adaptive systems. The latter strand takes up an agent- and rules of interaction-based approach to modeling complexity. The first explains the behavior of systems that can be modeled by complex equations where the output of one calculation is taken as the input for the next. These equations are repetitive and iterative.
Chaos theory explains how the parameters in the equations cause patterns in time. These patterns are called attractors. A parameter might be the flow of information or the amount of energy in the system. At low rates the system moves forward displaying a repetitive, stuck behavior. This pattern is called a point attractor. At higher rates the pattern changes. At very high rates of, for example information flow, the system displays a totally random behavior. The pattern is highly unstable. However, there is a level between repetition/stability and randomness/instability. This level is called the edge of chaos. The pattern in time is called a strange attractor. The strange thing with a strange attractor is that the ongoing movement is never the same but always recognizable. The pattern is paradoxically stable and unstable, predictable and unpredictable at the same time. These patterns are spatially called fractals.
Chaos describes a dynamic that is not a synthesis of order and disorder. It is about orderly disorder or disorderly order. The very meaning of these words is transformed.
The weather is normally used as an example of a system that displays this pattern. The overall weather patterns can be (almost) predicted over short periods of time. Over long periods, the behavior cannot be predicted. The long-term behavior of a system like this is determined as much by the smallest changes in the smallest of parts of the system, as it is determined by the laws governing it. The conclusion is very clear. Predictability is always short-term. Long-term predictions would only be possible if absolutely all the variables in the system could be measured with absolutely infinite accuracy. But it is impossible to know all the variables and it is totally impossible to measure all the variables with the accuracy needed.
The smallest overlooked variable or the most minute change can escalate up by non-linear iterations into a major transformative change in the later life of the system. Another conclusion is that from a chaos theory perspective, movement towards equilibrium is always movement towards death. If a system is healthy, successful and alive, it is “at the edge of chaos” where the long-term cannot be seen.
Classical physics took individual entities and their movement (trajectories) as the unit of analysis. Chaos theorists such as Ilya Prigogine, claimed that these trajectories cannot be calculated because of the impossibility of measuring with the precision needed. But there was something even far more exciting going on. Henri Poincaré was the first scientist to identify two distinct kinds of energy. The first was the (kinetic) energy in the movement of the particle itself. The second was the energy arising from the interaction between particles. When this second energy is not there, the system is in a state of non-dynamism. When there is interactive energy, the system is dynamic and capable of novelty and renewal. Interaction creates resonance between the particles. Resonance is the result of coupling the frequencies of particles leading to an increase in the amplitude of motion. Resonance makes it impossible to identify individual movement in interactive environments because the individual’s trajectory depends more on the resonance with others than on the kinetic energy contained by the individual itself.
Every interaction of any particles is thus potentially meaningful and can lead to amplification of the slightest variation. Interactive systems with even the smallest variations take on a life of their own that is under continuous construction. The future form and direction of the system is not visible in the system at any given time. The future is not in the system and it cannot be chosen or planned by anyone.
The scientists at the Santa Fe Institute developed the other strand of research: the complex adaptive systems approach. A CAS consists of a large number of agents. Each agent behaves according to its own intentions and rules for local interaction. Local here means that no agent can interact with the whole population of agents at the same time. No individual agent can determine the pattern of behavior that the system as a whole displays. These adaptive systems display the same dynamics as the chaos theorists found: stable equilibrium at one end of the spectrum, random chaos at the other, and in-between the newly found complex dynamic of stability and instability, predictability and unpredictability, paradoxically at the same time: the edge of chaos.
The conclusions are important for us. Firstly, novelty always emerges in a radically unpredictable way. Secondly, the patterns of healthy behavior are not caused by competitive selection or independent choices made by independent agents. Instead, what is happening, happens in interaction, not by chance or by choice, but as a result of the interaction itself.
The Internet changes the patterns of connectivity and makes possible new enriching variety in interaction. The changed dynamics we experience every day through social media have the very characteristics of the edge of chaos.
The sciences of complexity change our perspective and thinking. Perhaps, as a result we should, especially in management, focus more attention on what we are doing than what we should be doing. Following the thinking presented by the most advanced scientific researchers, the important question to answer is not what should happen in the future, but what is happening now?
Our focus should be on the communicative interaction creating the continuously developing pattern that is our life.
Thank you Stu Kauffman and W Brian Arthur. Based on Ralph Stacey and Doug Griffin.
Responsiveness, emergence and self-organization
March 21, 2010
I gave keynote speeches at two conferences this week. The organizers of the events did not suggest a (#) hashtag to be used by the delegates. There wasn’t any backchannel Twitter discussion going on in the audience. I felt strange.
I wasn’t able to listen and respond to real-time feedback. I was missing the self-regulation and self-organizing that social media make possible. This is what I have grown so accustomed to. I started to ponder on two questions: Is it becoming more common for responsiveness to be the missing ingredient in many communities? And can there be rules for responsiveness that help to create viable communities?
I know that there are problems with two-way communication. There are the people with a pre-set interpretative model. We all know the people who are grinding their axe at the back of the room. They are the know-alls and the one-point-of-view evangelists, the people who insist on bringing all conversations round to their particular issue.
I know that there are even bigger issues: All participants are never visible. Any given conversation on the Web may have a few active participants and several silent ones. This creates a fundamental imbalance in the system and gives the oddballs the opportunity to dominate the space in a way that would be much harder to do off-line.
What I felt at the conferences was a crucial disparity: they hear me talking, but I don’t hear them. The audience was both present and absent at the same time. A conference with a Twitter backchannel creates inputs from the official speakers and responses coming from the audience that is present, but also the online audiences elsewhere. The most important thing is that the primary inputs can then be further adjusted on the basis of the responses from the group. There is real-time emergent self-organizing going on.
Information flows are far too often unidirectional. The audience is present but in a passive, invisible way. The tyranny of the hatemonger results from this one-way flow and scarcity of feedback.
The volume is too high for any single individual to filter out the useless or plain repulsive. There are, however, ways to filter out the irrelevant and the obnoxious, but it requires people to respond. If you are a participant, you are also a moderator.
The quality control has to be handed to the community itself without any single individual being in control. The solution is fairly simple in theory. It is about responsiveness and a mix of negative and positive feedback.
You always rate what you see. The ratings coalesce algorithmically into something that is called karma in Slashdot. If your contributions are highly rated you get karma points. The karma you have earned means that your subsequent posts begin life at a higher level than posts by others. Your ratings also have a higher value than ratings given by people with fewer karma points. Dynamic rating is to posts what links are to websites.
The people worth following, the leaders, raise bottom up. Hierarchies in network architectures are natural and dynamic heterarchies. In fact this is the only way that there can be leaders in democratic systems.
One algorithm tracks the value of contributions; the other tracks the value of contributors.
The Web 2.0 gave the audience a voice. What is happening at the moment is much more radical. It is not about representation but gestures and responses leading to emergence and self-organization. It is not about the message or the media any more. It is more about the rules of responsiveness. In a simplified way, you can express those rules as positive and negative feedback moving the whole system towards a particular direction based on the behaviour of the participants.
The definition of what is quality and what is crap is a result of the responsive interaction. It is not groupthink however, because the ratings of people with high karma points weigh more than the assessments of the average members. The huge problem is that the majority viewpoints get amplified, while minority opinions get silenced.
This is why we need a new category to support quality. It is diversity.
Changing the algorithm to reward diversity of opinion means the emergence of a system that looks totally different. Instead of highlighting posts with high average ratings, the system could highlight posts that have triggered a high divergence of ratings. There are many +5 responses, but also many -5 responses. The posts that inspire strong responses either way, both positive and negative, could then rise to higher visibility. The system can thus reward controversial voices, not only popular ones.
A viable system needs to reward perspectives that deviate from the mainstream.
We need perspectives that don’t aim to please everyone. The oddballs would still be marginalized but the thoughtful minorities who attract both admirers and critics would have a visible place in the ongoing process of creating the future in responsive collaboration.
Thank you Steven Johnson
Lessons from the early adopters of corporate social media
March 15, 2010
The mainstream approach to management places a heavy emphasis on the formulation of plans and intentions and then communicating them as actions to be implemented by the organization. The starting point for change involves conceiving a picture of the future that is somewhat different from the picture of the present. After the content side is taken care of, the focus is then on providing tools for the process of change.
The approach that is made possible through enterprise social media is very different. The question that is now asked is: “How can people participate in such a way that things develop and change over time?”
The strategic focus of the early adopters of corporate social media is an ongoing continuous movement that is open-ended, and always incomplete. The strategic logic has been temporal rather than spatial. When following a spatial metaphor, there is a territory that can be explored and understood, but here the territory is seen as being under continuous development and formation by the exploration itself. “It is impossible to map an area that changes with every step the explorer takes.” People inhabit a world of emergence, uncertainty and responsive change.
Themes such as communities, social network analysis and social graph underline a fairly strong sense of definable relationships and a sense of “us”. Our studies, however, show that social media create a dynamic and shifting sense of groups one belongs to. Conversations always follow from previous conversations and move on involving others, often as a result of responses from outside the corporate firewall. Work utilizing social media has much less clear and managed beginnings and endings. There is, typically, no pre-conceived design for the pattern of work: it evolves live.
Corporate life is improvising together
Physical meetings in organizations are often more or less orchestrated and planned in advance: “You should come prepared. There should be a clear goal for the meeting.” Following this thinking, there is no true sense of creating the future together. It is much more likely that people construct what they have always constructed. When people use social media to connect, they experience the potential inherent in communication, depending on how they express themselves, and how they respond. “Social media create the experience of acting into the unknown, creating the future together, improvising together.”
By linking improvisation to a group, like in theatrical improvisation, we get to what is in fact happening in social media. All of us with our differing intentions, hopes and fears, are acting in corporate plays that are very close to improvisational theater. We are self-organizing in shifting social configurations in the responsive interplay of different players.
We are fellow-improvisers in corporate ensembles constantly constructing the future, and our part in what is happening, in responsive interaction. The idea of improvisation is often associated with notions of unrehearsed, unintentional action. However, the more skilled we are, the better we can improvise. The better we have planned, the more flexible we can be. The more intensely we are present, the more responsive we can be.
The real time web is creating a real time company
The most important outcome is that social media focus attention more on what people are doing in the present than on what they intend to do in the future. The focus is on communicative interaction, the next tweet and the latest blog post.
The pattern of relating also becomes very clear: “We get to see who is talking and who is silent? Who is invited to join and who is excluded or opts out?” The focus of attention is on the processes of participation and the life stream as the narrative of progress.
A senior manager in a very large multinational corporation explained the impact of social media: “Since I moved away from thinking that what I do is manage the corporation through communicating with the whole corporation, I have started to pay attention to my own participation with the people I meet or should meet, and my responses in everyday interaction. Through asking different kinds of questions and through pointing to different kinds of issues, through changing my own participation, I have in fact changed my company.”
Thank you Keith Johnstone, Srikumar Rao, Patricia Shaw and Doug Griffin
Interactive value creation
December 4, 2009
The division of labor reduces organizational effort and the cost of work. The division of labor also increases the quality of efforts through specialization. For this reason all societies and all enterprises are heading, at least to some extent, towards specialism. The assumption has been that the further the division is carried, the greater are the savings and the better the quality of the contributions. This has led managers to focus on the efficiency of activities separated from other activities and organizational design and management are seen as the planning and execution of a collection of independent activities forming the organizational system.
The function of the line manager was accordingly to be the representative of his box, his domain of action and resources. The manager enjoyed a high degree of autonomy and was accountable only for that domain. The grounding principle in practice was: “Don’t tread on my grass, and I won’t tread on yours”.
From action to interaction
As demands for higher value and creativity are the norm today and the complexity of offerings has grown, we have begun to see that the division of labor has reached its point of diminishing returns. What managers have learnt is that the division of labor always implies a scheme of interaction by which the different divided activities are made to work together. The lines between the boxes are starting to matter more than the boxes! Complex value creation is impossible without interaction. This is because any higher-value activity involves complementary, often parallel, contributions from more than one person or one team. In fact, the more complex the offering is and the more specialized the resources needed, the greater the demand for the amount, quality and efficiency of communication, because of the inherent interdependence of the activities.
One-dimensional approaches to interaction have involved top-down command-and-control or sequential workflow-based communication, where the action of one part is meant to set off the action of another. Interaction has thus been seen as one-way signals, a system of senders and receivers (Shannon and Weaver 1948). These approaches seemed to work in simple, low-value environments, but are not creating the desired results any more. What managers have lately found out is that in the pursuit of higher value and when facing the growing demands of complex offerings the value of actions is limited by the value of the interaction. The two are mutually dependent
Activities and interaction are mutually dependent
A system of partial activities that go into the completion of the total offering always implies a scheme of interaction among the persons concerned. If the scheme of activities changes, even somewhat, the scheme of interaction should change too. As the two are mutually dependent, it means accordingly that if there are changes in interaction, so the activities will change.
The mainstream management paradigm is based on the presupposition that activities are the independent, governing factors and the scheme of interaction conforms to the planned division of labor as a secondary feature. The organizational structure, as a number of independent activities, comes first. Then an appropriate system of co-ordination and communication is put into effect. If, however, action and interaction are mutually dependent, it means that low-quality interaction leads to activities that are poorer than planned, just as enriching, high-quality interaction may lead to higher-value activities than planned.
We need to understand how the present ways of dividing labor have been historically based on a very different communications environment than the one we are living in at present. The earlier high cost of coordination and communication is the reason behind many of the organizational forms that are taken for granted and which we still experience. The digital world we live in today is totally different when it comes to the transaction costs associated with coordination and communication and allows us to experiment with totally new value creation architectures.
Managing the scheme of interaction – creating a social media/communication strategy
The activity systems and units of activity can no longer be seen as a collection of independent activities and independent high-performing specialists. There are, however, many challenges ahead if we adopt the way of thinking of seeing interaction as the governing factor in organizations. One of the challenges is our language. That is the way we speak about work following the system of subjects and predicates. Our language of work is geared towards handling one independent factor and one dependent factor at a time: “someone is doing something to somebody”. Linear causes and effects, rather than thinking in terms of mutual interdependence and non-linearity, are built into our management speech. And yet, a situation that can be described accurately in terms of linear, rational causes and effects is the least common one in social contexts. An organization consisting of people is always a social network following a different logic – complex causality. Organizations as social activity processes are about interdependent people working in complex interaction.
If we take this view, it means that people and actions are simultaneously forming and being formed by each other at the same time, all the time, in interaction. Instead of thinking in terms of spatial metaphors, of organizational levels, boxes and lines, this explanation focuses attention on how the actions of people create patterns in time following a very different approach to communication than the sender receiver model.
Organizations seen as patterns of interaction
Organizations can be described as patterns of communicative interaction between interdependent individuals. All interacting imposes constraints on those relating, while at the same time enabling those people to do what they could not otherwise do. Supportive, inspirational, energizing and enabling patterns of interaction are the most important raison d´être of working and being together. If we see interaction as the governing factor and see organizations and organizing as relationships between interdependent people, our methods of sense making need to change. Social interaction does not follow linear causality, seen as a system of senders and receivers, but is fundamentally non-linear, responsive and complex. Following this logic, organizations today and information-based value creation in general can only be understood if seen as complex, communicative patterns of mainly digital interaction.
Back from interaction to action
Resource allocation has always been one of the main tasks of management: planning what is to be done by whom and by when. In integrated systems and with homogeneous resources, this allocation can easily be performed top-down and in advance. Planning can take place separately from action. When knowledge resources are the decisive factors of value creation and when work takes place in digital, global, decentralized environments, this top-down process is increasingly inefficient. A manager cannot know who knows best or where the most valuable contributions could come from. The solution has been so far to try to “know what we know”, and, even more importantly, try to “know who knows”. Neither of these approaches has quite fulfilled expectations. Knowledge databases have not met the situational needs of their users. Accordingly, people have not been able to explain what they know to others or even to themselves in a meaningful way
Because of the aforementioned growing needs in daily organizational life a new, different approach has to be adopted. One could even claim that a new mode of knowledge based production is now emerging in, and because of, the digital networked environments. The most important platforms for the new production systems are social media platforms.
This new production method refers to a new economic phenomenon: people from the whole network contribute pieces of their time and expertise to tasks, emergently, according to their interests, availability and experience, working in a transparent, open environment. This method has systemic advantages over traditional production hierarchies when the work in progress is mainly immaterial in nature and the capital investment involved can be distributed. For most knowledge-based products and services, this kind of production is the most efficient method of creating value from a resource allocation point of view.
The system is developed as much in a bottom-up manner as a top-down one. In a top-down system everything is created and provided by the organization to the user. The user has no or very little control over what services, information and people are available to him. Instead of forcing people into predetermined groups in the way groupware does, social media facilitate the natural formation of groups based on spontaneous, contextual needs for interaction. In social media, people affiliate through personal choice and need. Understanding this difference in community formation is crucial for building self-sustaining, dynamic communities.
A Wiki is a typical knowledge production medium, a platform for interdependent people to work in parallel interaction. A Wiki provides the most efficient way for a group of people to contribute, edit and interact with information that is meant to be shared. A Wiki can be seen as a way to create and iterate collective information, thus developing shared iterative learning. It’s about making visible what has been learnt and the road that leads to it. This leads to a better sharing of experiences, use of skills and utilization of the total number of brains in the network.
The primary goals are increasing the value and quality of information and the value and quality of interaction and at the same time lowering the transaction costs associated with information and interaction. Even more importantly, open interaction platforms such as Wikis are a medium for sharing what we would like to know next, where we would like to go, and what we would like to explore. A Wiki is a medium for continuous, creative learning. This thinking is based on a belief that everything can and should develop in iterative interaction among the network of users. In practice it means voicing questions and concerns for others to answer of their own free will: the small deviations, the small questions that we don’t normally pay much attention to or have time to explore. These are, however, the starting points for change, improvements and learning. There is a shift in thinking from sharing what we know to sharing what we don’t know.
All organizations essentially operate like Wikis. Every organization has it’s own language, resulting in a unique, iterative understanding of concepts, terms and ongoing sense making. There is always a lot of information that is continuously evolving in the “encyclopaedia” of an organization. The articles are things like strategies, customer databases, product information and manuals.
Interactive, iterative work
In these kinds of contexts, information artifacts that don’t connect with ongoing live conversations are often of less value, even obsolete and most probably out of date. Because of this, we are now moving away from a focus on content to a focus on conversations. Content should be seen as the by-product of conversation. Perhaps in the future of digital work IT will not mean Information Technologies, but Interaction Technologies.
This view focuses attention on the way in which daily, mainly digitally mediated communication between people organises value creation and, at the same time, creates value. An organisation should today be understood as complex, self-organising, iterative patterns of interaction, through which both continuity and innovation emerge as patterns in time.
