Self-synchronization and the System of Creation (SoC): a hidden order in Nature, Society, and our Minds

Mixel Kiemen
21 min readNov 5, 2023

--

This article goes back 20 years, how my research started as Artificial General Intelligence (AGI), through A-life experiments on creativity. It set me on an interdisciplinary PhD journey investigating the System of Creation (SoC), which eventually resulted in a meta-system explaining creation in modern society. 5 years ago the research returned to consciousness/psychology states in an unexpected way as personal development. In this article, the experience is integrated as the mesa-model. Trying to elaborate on the SoC has been challenging, a simple video on self-synchronization may help the introduction.

Take a minute to look at this video on self-synchronization. It shows five metronomes on a movable base. The metronomes start out of sync and it takes only a minute for all to sync. It is a beautiful and simple demonstration of self-organization. The technique can help explain the System of Creation (SoC) behind the “origin of X” research, where X is a variable. I began with AGI research on the “origin of creativity”, which was strongly influenced by the AGI research on the “origins of words”. I found many colleagues from other domains who would explicitly, or implicitly, work on an origin study. This reveals several SoC cases. It all started in the summer of 2002 with the research question: can we use A-life to demonstrate the origin of a creative action?

Before diving in, the AGI researched a short overview of this article. The first section will focus on the AGI research, mimicking nature and elaborating the default SoC model. The SoC model is an open-ended cognition model emphasizing an extended mind capable of developing science and technology. So we shift to the level of society. The next section is about the PhD research where five cases are recognized that have this SoC model as a hidden order. The cases will become combined into a meta-model, explaining a lot about the hidden dynamics behind the creation process in modern society. The meta-model is used to develop a system that in theory could resolve Grand (societal) Challenges.

Next to the meta-model, also infra-models are recognized in studies on “the origin of life”. Developmental biologists describe how problem-solving by organs (during embryogenesis) and cells (during gene expression) have their own workspace, which is the base of the SoC model. The infra-models, the default model, and the meta-model are about the same outward problem-solving challenges.

This article will not go into the infra-models (see this video for an intro or my PhD chapter 3). The infra-model is mentioned to emphasize the difference with the mesa-model. The mesa-model is an inward embodied experience revealing some wicked features of consciousness. The last SoC case at the level of Society (i.e. meta-model) will be the first to address an inward process with collective psychology. In contrast to the meta-system transition (outward), a mesa-system transition (inward) is recognized: from collective psychology to personal growth.

The mesa-model has psychological states and one state was used to explain the AGI research, so we return to see all other states. First, all states act unconsciously and the SoC dynamics can explain how the awareness arises. The article ends by considering evolution at work by the meta- and mesa-system transitions. The forces are actively changing society. Humanity is seen as the catalyst for the development of a global living-system.

Self-synchronization and Artificial General Intelligence research

My AGI research started in 2002–2003, with the two professors who founded the AI lab at the Vrije Universiteit Brussel. On the one hand, Luc Steels with the origins of words experiments done in 1999–2001. The experiment applies advanced self-synchronization. On the other hand, Walter Van de Velde. He became CTO of Starlab between 1996 -2001 (i.e. the dot-com period). Starlab was an experimental R&D facility for Media. Another founder of Starlab was Nicholas Negroponte (MIT Media Lab founder). So the lab had some strong scientific foundation, which strangely did not disseminate after the bankruptcy. Walter Van de Velde came back shortly (2002–2004) to his former lab and became my first advisor before leaving research as a policy officer with the EU Commission.

I turned the insights from Starlab experiments, and the origins of words, into an experiment on the origin of creativity. Practically it started with the A-life simulation mimicking the processes of creativity. A bit more insight into the origins of words is useful as an introduction. The experiment is about software agents who “talk” to each other and self-synchronize to one lexicon. So how does it work?

Look at the video on self-synchronization and let us analyze it. Notice the base moves as well. The base gives feedback to all metronomes and results in synchronization. This follows a simple principle of feedback. The origins of words follow a similar principle to self-synchronization to one lexicon. Each agent had a list of the same words and a list of the same meaning but at the start words and meanings are randomly connected. Each agent was like one metronome and talking to each other allowed self-synchronization. The words were all similar, forming the base, and talking would synchronize the means of each word.

The origin of creativity applies a similar self-organization technique, now focused on one agent interacting with a changing environment. The agent was more complex and had an internal base called the working memory where four processes function as self-synchronizing metronomes. Two of the processes help connect the three involved mediums: Long Term Memory (LTM), working memory, and environment. The two other processes contribute to a system of control able to be creative. So how did we get to this design?

At the beginning of the experiment, we had no idea what kind of system was needed and simply focused on a practical setting: the agent eating nuts. The agent would have the knowledge to survive in its environment, but its actions would alter the environment and the agents had to become creative. Using classic action research (plan, execute, reflect & repeat) the simulations removed simplifications to get closer and closer to the actual creative process.

The ambiguity turned out essential, ambiguity of perception, ambiguity of action, and ambiguity of meaning. Nutfruit would lie open on the ground, some partly still in the shell and others fully closed. Eating the most approachable nuts first would result in a journey where the agent shifts from: picking up the food, to wriggling out the fruit, to breaking the shell to get to the fruit. At the same time, the agent looks different to the environment: from searching for brown fruit, to searching for the green shells. To end up with very closed shells that cannot be opened (i.e.perturbations), which gets solved by knowledge exaptation. Let me elaborate on this last claim with a simple explanation and support it with a simple drawing.

The agent knows there is food in the nut and cannot reach it (perturbation). The agent starts an internal thought process finding associations from different contexts to create an experimental action never tried before (exaptation): using a stone to break the shell. This action is never done before and would lead to an experience never received before (i.e. “novelty”). For details on how this translates to code, see my PhD second chapter 2.

Analyzing the experiment led me to understand how ambiguity and embodiment are essential. Of course, this has little relation to machine learning (i.e. basic AI). To continue the research I expanded my horizon by going to conferences on cognition and so I learned about the Global Neuronal Workspace (GNW) Hypothesis (left figure), which is almost identical to the SoC model (right figure). Both describe four similar input processes in a workspace and one output process. Let me create the link between the two models:

  • (Internalizing) Long-Term Memory: the past, holds all kinds of associations
  • (Externalizing) Perceptual system: present, brought in the workspace by tags
  • (Directing) Attentional system: focusing on challenges
  • (Evolving) Evaluative system: learning from experience

The GNW hypothesis also has an output process to the Motor system, showing more about the embodiment of the SoC. In the SoC, this is a bit more complicated as the evolving process has its input to the workspace via the external dotted lines. The output goes in two directions (dashed line) allowing the agent to master the environment or model the LTM.

Notice that the LTM became the “system core” in the SoC model, and working memory became the workspace, this is a generalization where other SoC cases will not be about memory. The four input processes go outward by solving problems, while the output process goes inward by embodied experiences. The inward going will relate to psychological states, by the end of this article I will elaborate on how this happens via a mesa-model. To demonstrate how the motor system is involved with cognition a deeper dive into the Gestalt state of perception is given:

The left part of the figure is what can be measured, the rapid eye movement, while looking at a picture. To show this clearly, the rapid eye movement is projected back onto the image being perceived (second picture). The right sketch demonstrates the two synchronizing processes: Internal (LTM, past) and External (Perceptual, present). The self-synchronization happens as follows:

You open your eyes and see dark / light (no shape yet). In the sketch, this is represented just below the word external, as a vague shape. The light triggers the externalizing process to iterate on the boundaries. At the same time, we have no clue what is happening internally (question mark in the sketch). The eye will sharpen the boundaries between light and dark to find a round shape. The tag “round” is put in the workspace and the internalizing will find several associations: bal, apple, face. The associations are pushed into the workspace, each containing specific tags The externalizing can see if such tags exist in the environment.

The dynamic happens unconsciously in milliseconds. One way to verify this dynamic is happening is by considering cases to trick the dynamic by Change blindness experiment. Internalizing and externalizing simply synchronize the three mediums: LTM, workspace, and environment. The control comes from directing and evolving. For example, the face could trigger all kinds of challenges, at first still simple: maybe you want to talk to this person, maybe you want to avoid the person. More complicated challenges will follow, requiring ever more complicated associations and tags.

The evolving process is always active and by default a simple conditioning process (i.e. learning by neural networks). One particular challenge can be to master a new skill, now the focus is on the environment and most internal thoughts are filtered out. Similarly one can go inward by modeling the LTM and neglect signals from the environment. Expressing the SoC dynamics a bit formally as a control system was done in chapter 4 of the PhD.

With mastering and modeling the environment can become an extended mind i.e. creating science and technology. Investigating the extended mind will result in a journey and I discover many SoC cases along the way. Eventually, the meta-model is recognized, explaining a hidden dynamic in modern society. Allow me to give a chronological overview of the journey in the next section.

Meta-system transition, turning the research from individual to society

The A-life simulations were done in 2002–2003 and analyzed in 2003–2004. The next year only minor improvements happened to the SoC model. I was happy to re-discover SoC in the GNW hypothesis. Clearly, I was on to something! Little did I know how fundamental the discovery would be. It was strange to find non-cognitive cases of the SoC model by examining the extended mind. It seems like the pattern is this: first people create an artifact/model in a very protected environment, next they enhance the model to work in complex adaptive environments and create another SoC case.

The first non-cognitive SoC case was a system described by science studies. Science studies have been looking at Science in Action (SiA) and in one particular book they ask “What is the origin of science”? This book was the biggest reason to shift from AGI research to cybernetic research, as the SoC was not limited to the domain of cognition (Kiemen 2008). The third case I found was by participating in innovation management communities that referred me to organizational strategy research describing the Dynamic Capability (DC) framework. We could express the research as “the essence/origin of the corporate organization”.

In the domain of methodologies, another SoC was found: the Action Design Research (ADR) method (2011). The ADR method is a special SoC case in two ways. First, as a model about a method, it goes into a recursion. We could express it as “the origin of creating origins”. Second, as the fourth SoC case, I would realize a meta-model would be possible: GNW hypothesis (internalizing), Science in Action (externalizing), DC framework (directing), and ADR Method (Evolving).

While the insight came quite suddenly, it would take me years to refine the meta-model, with a whole network of supervisors. We called it the Agile-Enterprise Innovation Planning (ÆIP) system (Kiemen 2015, p.183) explaining how modern society creates itself. The SoC cases are used as foundations and transformed to become the four processes of the meta-model, they are:

  • (Internalizing) Cohering model: making sense collectively (narratives)
  • (Externalizing) Eventuating model: making the world (science & tech)
  • (Directing) Strategizing model: making corporate organizations
  • (Evolving) Establishing model: making rules of engagement (method)

Validating the ÆIP system as a whole was a bridge too far for a PhD study. So we only experimented with the Cohering model. The experiment was done in 2011–2012 involving 400 students in a semester course, using a platform for collective education (i.e. sensemaking). After my PhD (2015–2016) I repeated the sensemaking applied to SMEs innovation, which I still hope will become an academic publication soon (so no future details, sorry). Today my postdoc work is mostly focused on the Eventuating model with small progress on Strategizing and Establishing models.

The development between 2017–2020 is more related to the mesa-model and a bit disconnected from the meta-model. Allow me to jump to 2020 and the discovery of the SoC case that will function as the output process (i.e. turning inward) of the ÆIP system. The ÆIP system is about societal innovation and it needs to be embodied in actual organizations with actual teams doing actual work. This is very different from the Strategizing model. The last case was found by participating in communities on sociocracy i.e. the Sociocratic model.

Both the Strategizing model and the Sociocratic model are about organizations, the first goes outward, and the second goes inward. Outward organizations are asked to deliver value for their stakeholders, creating lean processes and organizational capabilities. Inwards organizations are asked to collaborate and govern, creating agile processes and organizational identity. The Sociocracy 3.0 (S3) community has created artifacts (under Creative Commons licensing) to help going inward and one poster describes another SoC case:

  • (Internalizing) Responsiveness: structure for collective intelligence
  • (Externalizing) Participation: now about the collective efforts of an organization
  • (Directing) Doing: prioritizing and pulling work (essence of directing).
  • (Evolving) Agreeing: experiment and learn, but again in a collective way.

In the GNW hypothesis, the rationality developed in the agent’s workspace goes inwards into the body, creating an embodied experience. The ÆIP model elaborates on how society creates artifacts and they can go inwards into organizations. Some examples: In commercial organizations, these are products and services. In social organizations, artifacts can relate to security, healthcare, education, etc. In religious organizations, artifacts have spiritual value.

Today the ÆIP model is theory and the meta-model an evolutionary force. With society needing to resolve grand challenges, pressure exists to make the ÆIP model more explicit. This pressure is what allows my research to continue, now mostly on the Eventuating model, creating R&D that resolves the climate crisis. The other ÆIP model may be applied with a systematic solution to Grand Challenges, which is emerging by the Next Generation University. Considering how long it took me to get this far, it seems decades will be needed to see the full development.

Communities of professional coaches working on effective organizations (like S3) show us how this inward process has a strong relation to collective psychology, psychological safety, artful communication, etc. So we finally got to psychology but by the collective. Just as the infra-models show how all intelligence is collective, it appears all psychology is collective too. The particle-wave duality from quantum physics may be a good metaphor for understanding the personal-collective duality of cognition. So we have come full cycle: from the basic AGI research to the extended mind, to collective psychology, to personal development and finally advanced AGI with the mesa-model.

Mesa-model and the direct experience of consciousness

The mesa-model is about an inward-going process: the direct experience that everything is consciousness (Spira 2021). Let me begin with a metaphor to explain the research between 2002 and 2016 as an unsuccessful quest for consciousness. A sudden shift in the reference frame would explain how absurd the search was to start with:

Consider you are driving a car to find yourself. All the time you look through the front window. You drive everywhere to find yourself and never do. At night, when you finally put the car to a halt and turn inwards by switching on the light in the car, the front window turns into a mirror. There you are!

The SoC research started with a question about “real” intelligence/consciousness. Focusing on problem-solving skills was outward going. The inward going was implicit, as action research (creating artifacts) and participation research (in communities), which do contain the psychological states of the mesa-model (at the collective level). My “mirror moment” of consciousness is the moment the shift occurs from collective psychology to personal experience of the effervescence state. Now I realize the mesa-model was active all the time and contains four psychological states:

  • (Internalizing) Gestalt state: the way things are put together (content)
  • (Externalizing) Presence state: the ability to be aware (direct)
  • (Directing) Flow state: balance between skills and challenges (transition)
  • (Evolving) Effervescence state: creating spirit (transformation).

Each state reveals an aspect of consciousness. The Gestalt state is the content of consciousness. The presence is the direct experience of consciousness. Flow is the transition of consciousness (via experience & learning) and effervescence is the transformation of consciousness (via personal growth). The last state will become the hardest to explain.

Gestalt state and Presence state were demonstrated above with the motor-control of perception. Presence is about being there (Clark 1998): putting the brain (i.e. workspace), the body (i.e. LTM), and the world (i.e. environment) together again. It is important to know the mesa-model refers to the Gestalt state and not Gestalt psychology. The latter is a method of therapy, which got a lot of criticism. Yet the qualia experience of the Gestalt state is not being questioned.

Being there can be directed external to the world (extended mind) or internal to stories (imaginative mind). External is open-ended to the meta-model and internal is open-ended to the mesa-model. Adding the flow state is adding emotions and learning: boredom if skills are high and challenges low, anxiety if skills are low and challenges high. If both are balanced we feel truly alive.

While flow is a physical process commonly known, effervescence is a less known chemical process: the escape of gas from an aqueous solution e.g. the bubbles in drinks. During the AGI research effervescence was present (in spirit/mind) as the eureka effect, which was the focus of the simulation. To understand this state, we should go to Durkheim (1912) who introduced collective effervescence in his research on the origin of religion. He identified effervescence as the experience of the sacred in the group activity and elaborated on the essence of rituals.

Cognition has this personal-collective duality (metaphorically like the particle-wave duality). Triggering the effervescence state will be about the amount of experience: be it distributed in the collective or internalized in a person. To consider how personal development is internalized in a person, we can introduce Kegan’s (1995) five stages:

The first is the impulsive mind, i.e. doing without thought. The second, which all adolescents reach is the imperial mind i.e. very egocentric. 6% of adults never grow out of the imperial mind, so 94% reach a socialised mind. The next stage — the self-authoring mind — is only reached by 36% and only 1% reach the self-transforming mind (see drawing). From personal experience, I know this is fractured, based on interest. For example, some professors may understand all paradoxes in science and still be stuck socially with an imperial mind.

The state of effervescence creates personal growth and these stages can relate to different domains. Participating in many communities shows each time how personal growth exists in the domain and is not developed outside the domain. Allow me to add more examples next to the professor. A social worker may be able to embrace paradoxes about personal growth and have an imperial mind towards authority (like science and politics). A politician may embrace the paradox of running society and have an imperial mind to logic.

Now my personal development, concerning religion, I had an imperial mind before 2017. Afterward, I integrated many religious community experiences and the work of Durkheim, which allowed me to leapfrog and approach religion by design science. The dynamics of SoC imply effervescence and flow can create a state of mastering (i.e. focused on presence) or modeling (i.e. focused on Gestalt).

Mastering & modeling for the extended mind is about separation (i.e. individuation), for the imaginative mind it is about unification. To give a non-religious example with collective effervescence, consider the ecstatic state with festivals or big sports events, allowing the participants to unify as a community (or a nation).

Small amounts of mastering and modeling are required to stay in a conscious state, or things go on autopilot. You may have experienced this while driving the same road multiple times. After many iterations, it can happen you step out of the car and cannot remember driving there. The bigger the buildup the clearer the effervescence state. Like the spiritual master meditating for 30 years on a mountain. Or a person in a life-threatening situation, suddenly gaining this spiritual awakening.

The grade of experiencing effervescence is: feeling, becoming drunk of it, and reaching the sacred. Most people know this grade from alcohol, although it may be the worst of all cases. Drinking one glass will trigger the feeling, drinking more gets you drunk and losing your inhibitor. Taking a “heroic dose” will make you blackout. Alcohol is the worst, other drugs, like psychedelics, are more interesting. Microdosing will give you the feeling, taking a bigger dose will get you drunk on unification (love), and going to a heroic dose can get you into a sacred vision. The transition doesn’t always need to be chemically induced, hard labor can get you in the state as well. The best-known example is Runners High creating an interesting state of presence.

Exceptionally this grade can also be reached by scientific work, as expressed by Fritjof Capra (1975) in The Tao of Physics. In my own experience after 15 years of mastering SoC (and being in life-threatening situations), I first became drunk with insight to eventually gain experience of the sacred: a vision/memory of the cosmos. It was an experience so profound and unexpected that I needed a spiritual sabbatical in 2017 to digest the experience. It took me another 5 years to finally integrate the experience into my research.

The mastering was by shifting outward, from the default model to the meta-model, which follows the concept of meta-system transition: interaction, coordination, and control. I now also see an inward-shifting mesa-system transition: from spirit/mind, to culture/collective, to sacred/personal. The mesa-system transition existed first unconsciously in the research. All psychological states were present in spirit/mind with the initial AGI research. During the participation research, they could be observed in the culture of the collective. With my mirror moment, it became unified with my personal growth.

For the past 5 years, I researched the inward-going development, so I can now express my thought on it. Considering how SoC is also non-human, implies the mesa-system transition can also be applied to science & technology: virtual (in theory), artificial (in labs), and unified (in society). I could go over every model in the ÆIP system to see how the mesa-system transition acts on the model. Yet this article is already too long. Let me wrap up by giving a vision of how all recursion creates a vision for society.

Recursive and fractalized evolution of society: the origin of a stellar species.

The meta- and mesa-system transition create outward- and inward recursions. The outward-going recursion on personal development is easy enough to understand. To quote Marshall McLuhan(1967): “We shape our tools and, thereafter, our tools shape us.” So the next generation will have the mind to create new organizations, technology, etc, which create new generations.

The inward-going recursion is a bit tricky. Because of the fractured personal development in different domains, it is hard to see all the latest layers together. For example, colleagues on collective effervescence, try to add a layer to social development by turning festivals into a stage for mass transformation. As applied in Israel, to unify different religious communities. The irony is deep with the recent Hamas massacre at the Supernova music festival.

The trend to focus on festivals is also seen in the built environment, which sees festivals as an opportunity to scale up living labs to the city level. Yet, the essential technology at festivals dates back to mass-media technology. So not the last technological layer. Therefore social development has not considered the unique collective effervescence event by technology amplification: the hive-mind experience.

Digital tech led to intelligence amplification, but these communities are not the most socially and politically advanced and we witnessed the collapse of the community spirit once the bigger world of corporate business entered the ecosystem. Just before the collapse, I was able to interview all entrepreneurs. I identified an extra policy that triggered the self-organization of innovation (Kiemen 2011) to see it collapse in the next years.

Recent digital tech sometimes gets very quickly adopted and almost as quickly collapses. For example the Clubhouse app during the COVID lockdown, the numbers of users were: 108k in November 2020, 2.7 million in February 2021, and back to 108k in September 2021. The people who were present in February 2021 may have found that unique event where artists and geeks came together to create collective effervescence at a global scale. Because of the short time, not a lot of dissemination was possible. The digital tech led to a scale and speed correction for the emerging Global Brain Embodiment (GBE).

The GBE scale correction is to not see people as neurons, but communities the size of cities as neurons. The current focus on festivals as pop-up cities is perfectly aligned with this vision. I also suggest a speed correction by replacing the metaphor of a technological singularity (going to infinity) with a huge S-curve and humanity now closing in on the midpoint. The speed correction would imply a slowdown, which relates to the collapse seen in 2011 and is even clearer in 2021 with the Clubhouse app. To have an impact, the self-organizing innovation needs to be longer and this will require a slowdown of the speed.

Where I see a huge technological S-curve, I do see a divine singularity coming up and it is seen in recent adoption of AI tools going even faster and they are not collapsing. Some do work on psychology, like illustrating the symptoms of anxiety. The relationship between tech, dreams, and psychedelic experience is dissolving into a sacred digital experience, helping people to release anxiety. In the hive-mind article, I also argue how higher scales could be reached by integrating AI.

I believe a higher scale hive-mind is required for Next Generation University (NGU) to metaphorically function as the nucleus of a region and do like how my current university, is in a high-tech region called Brainport Eindhoven, which aligns nicely with this vision. Trying to understand the meta- and mesa-system transition in this process improves our understanding of how evolution at work is giving birth to stellar species we could see as gods.

It seems no coincidence the planets have names of gods. The story starts first in science with the Gaia hypothesis: how the world has a global metabolism by biodiversity. Today with the mesa-model I’m a step closer to understanding the evolution of Gaia’s mind and our role in it. The mind of Gaia is forming at the same time humanity is becoming a multi-planet species and at the same time we gain cosmic perception by seeing gravitation (via Gravitational-Wave Observatory).

Humanity is at the frontier of building a stellar species, just like 4 billion years ago, ribosomes built the first cell. The infra-model and research on the origin of life may help enrich the metaphor: problem-solving by cells (during gene expression) and organs (during embryogenesis), where cells are cities and organs are nations. All this development is going to take millennials, which is from an evolutionary perspective fast.

The biggest question still puzzling me is if humanity is the catalyst or the fuel in the process. As a catalyst, humanity would keep having its place, like the ribosomes in a cell. As a fuel, we may transform (effervescence) and get extinct as so many ancestrial species did before us. Both are okay, the mesa-system transition teaches us we are a cosmic dream after all.

--

--

Mixel Kiemen

Research Logbook, on the general System of Creation (SoC) and concrete implementations like Next Generation University (NGU)